WO2017114017A1

WO2017114017A1 - Computer device, system and method for implementing load balancing

Info

Publication number: WO2017114017A1
Application number: PCT/CN2016/105945
Authority: WO
Inventors: 黄登辉; 康达祥
Original assignee: 华为技术有限公司
Priority date: 2015-12-30
Filing date: 2016-11-15
Publication date: 2017-07-06
Also published as: CN107078969B; US10171567B2; CN110113441B; CN110113441A; JP6544872B2; EP3316532A4; CN107078969A; EP3316532A1; JP2018523932A; US20180332105A1; EP3316532B1

Abstract

Disclosed are a computer device, a system and a method for load balancing, solving the problems in the art of service interruption caused by the switching of master and backup load balancing modes, and congestion caused by centralized load balancing. By creating a load balancing virtual machine in each computer device, when load balancing processing needs to be performed on a service of the virtual machine in the computer device, a relevant service message is only sent to the load balancing virtual machine in the computer device for load balancing processing, thereby avoiding a processing delay caused by congestion of centralized load balancing. Moreover, when the computer device fails, the virtual machine in the computer device also goes offline, without generating a demand for load balancing. Therefore, in a load balancing virtual machine fault caused by a single computer device fault, there is no service interruption brought about by master-backup switching caused by the load balancing virtual machine fault.

Description

Computer device, system and method for realizing load balancing

Technical field

The present invention relates to information technology, and more particularly to computer equipment, systems and methods for implementing load balancing.

Background technique

Load balancing is a collection of servers in a symmetric manner by multiple servers. Each server has an equivalent status and can provide services separately without the assistance of other servers. Through some load sharing technology, externally sent requests are evenly distributed to a server in a symmetric structure, and the server receiving the request independently responds to the client's request. The balanced load distributes client requests evenly to the server array, providing fast access to critical data and addressing a large number of concurrent access services. This clustering technology can achieve near-mainframe performance with minimal investment.

With the development of cloud computing technology, various services have migrated to the cloud environment, and the demand for cloud load has also been put forward for the load balancer. The virtual load balancer (vLB) has emerged.

At present, most of the virtualized load balancers are virtualized and softwareized by traditional load balancers, which cannot meet the requirements of cloud computing for high reliability, horizontal expansion and high bandwidth. To improve the reliability of the vLB service, the two vLBs are used to provide services in active/standby relationship. The heartbeat information is transmitted between the active and standby vLBs through the Virtual Router Redundancy Protocol (VRRP). When the standby device finds that the host is down, the standby device takes over the host and continues to work. The active/standby relationship implemented by VRRP takes place after the host is down. The active/standby switchover takes a certain amount of time, causing traffic interruption during the handover process. Moreover, all load balancing services must perform load balancing to a load balancing node, when load balancing access When the volume is large, the bandwidth of the centralized load balancing node becomes a bottleneck, causing congestion and delay of the service.

Summary of the invention

The embodiments of the present invention provide a computer device, a system, and a method for implementing load balancing to solve the problem of traffic interruption caused by handover in the active/standby load balancing mode and congestion caused by centralized load balancing in the prior art.

In one aspect, an embodiment of the present invention provides a method for implementing load balancing, which is applied to a computer system including at least two computer devices, where each computer device includes a central processing unit and a memory, and the at least two computer devices are included. The first computer device is a cloud management platform, and the second computer device is a virtual machine that initiates a service, where the service is a load balancing service, including:

The first computer device obtains configuration information of a load balancer to be created in the second computer device, the configuration information of the load balancer includes an identifier of the load balancer, and a virtualized IP of the load balancer Address VIP, the VIP of the load balancer indicates the load balancing service;

The first computer device creates the load balancer on the second computer device according to the configuration information;

The first computer device configures a forwarding mode of the load balancing service on the second computer device, the forwarding mode causing the load balancing initiated by the virtual machine initiating the service on the second computer device The service packet of the service is forwarded to the load balancer.

The virtual machine that initiates the service on the second computer device sends the service packet to the load balancer according to the forwarding mode;

After receiving the service packet, the load balancer on the second computer device selects at least one backend server to execute the service, and the backend server is configured to execute the load balancing service.

In a possible design, the first computer device is on the second computer device The forwarding mode of the load balancing service is as follows:

The first computer device configures a virtual switch deployed on the second computer device, and sends a MAC address of the load balancer to the virtual machine that initiates the service;

The sending the service packet to the load balancer according to the forwarding mode of the virtual machine that initiates the service on the second computer device includes:

The virtual machine that sends the service sends the MAC address of the load balancer as the service packet of the destination address, and the virtual switch sends the service packet to the load according to the destination address of the service packet. Equalizer.

In a possible design, before the virtual machine that initiates the service sends the MAC address of the load balancer as the service message of the destination address, the method further includes:

The virtual machine that initiates the service sends a control packet to the virtual switch, where the control packet carries the VIP;

The virtual switch sends a response message of the control packet to the virtual machine that initiates the service, where the response message carries the MAC address of the load balancer.

In a possible design, the control message is an address resolution protocol ARP message or an Internet Control Message Protocol (ICMP) message.

In a possible design, the first computer device configures a forwarding mode of the load balancing service on the second computer device, including:

The first computer device instructs the load balancer to send a MAC address of the load balancer to the virtual machine that initiates the service; or

The first computer device configures a MAC address of the load balancer on the virtual machine that initiates the service;

Correspondingly, the sending, by the virtual machine of the initiating service on the second computer device, the sending the service packet to the load balancer according to the forwarding mode comprises:

The virtual machine that initiates the service sends the MAC address of the load balancer as the service packet of the destination address, and the virtual switch deployed on the second computer device is configured according to the service The destination address of the packet sends the service packet to the load balancer.

In a possible design, the creating, by the first computer device, the load balancer on the second computer device according to the configuration information includes:

After the first computer device creates the first virtual machine that initiates the load balancing service on the second computer device, the load equalizer is created.

In a possible design, the method further includes:

And deleting, by the first computer device, the created load balancer after the last one created by the second computer device initiates the load balancing service is deleted.

In one possible design, the method further includes:

The selected backend server directly returns the result of executing the service to the virtual machine that initiates the service; or

The selected backend server returns an execution result of the service to the load balancer, and the load balancer returns the execution result to the virtual machine that initiates the service.

In another aspect, an embodiment of the present invention provides a method for implementing load balancing, which is applied to a computer system including at least three computer devices, wherein each computer device includes a central processing unit and a memory, and the at least three computer devices The first computer device includes a first computer device as a cloud management platform, and the computer device other than the first computer device of the at least three computer devices operates as a business computer device on each business computer device There is a virtual machine that initiates a service, and the service is a load balancing service, including:

The first computer device respectively obtains configuration information of a load balancer to be created on each service computer device, where configuration information of the load balancer includes an identifier of the load balancer, and virtualization of the load balancer IP address VIP, the VIP of the load balancer indicates the load balancing service;

The first computer device creates a local load balancer of each of the service computer devices on each of the service computer devices according to the configuration information of the load balancer to be created in each of the service computer devices;

Configuring, by the first computer device, a forwarding mode of the load balancing service on each of the service computer devices, the forwarding mode causing the virtual machine initiated by the originating service on each of the service computer devices to initiate the The service packet of the load balancing service is forwarded to the local load balancer, so that the local load balancer on each service computer device initiates a service packet to the virtual machine that initiates the service on the service computer device. Perform load balancing processing.

In a possible design, the forwarding mode of the first computer device configuring the load balancing service on each of the service computer devices includes:

The first computer device configures the virtual switch deployed on each of the service computer devices to send the MAC address of the local load balancer to the local virtual machine that initiates the service.

In a possible design, the first computer device configures a virtual switch deployed on each of the service computer devices, and sends the MAC address of the local load balancer to the local virtual machine of the originating service. It also includes:

The virtual machine that initiates the service on each of the service computer devices sends a control packet to the local virtual switch, and the control packet carries the VIP;

The local virtual switch on each of the service computer devices sends a response message of the control packet to the local virtual machine that initiates the service, where the response message carries the MAC address of the local load balancer.

In a possible design, the first computer device, before creating the load balancer on each of the service computer devices, according to the configuration information, further includes:

The first computer device learns that the first virtual machine of the first service computer device is created on the first service computer device, and the load balancer of the first service computer device is created on the first service computer device until A local load balancer on each of the business computer devices is created on each business computer device.

In one possible design, the load balancer includes a load balancing virtual machine, a load balancing container, or a load balancing namespace.

In another aspect, an embodiment of the present invention provides a method for implementing load balancing, which is applied to a computer system including at least two computer devices, wherein each computer device includes a central processing unit and a memory, and the at least two computer devices The first computer device and the second computer device are included, the first computer device is a cloud management platform, and the second computer device is running a virtual machine that initiates a service, where the service is a load balancing service, including:

The first computer device configures a forwarding mode of the load balancing service on the second computer device, the forwarding mode causing the load balancing initiated by the virtual machine initiating the service on the second computer device The service packet of the service is forwarded to the load balancer, so that the load balancer selects the backend server that executes the service after receiving the service packet.

The first computer device configures a virtual switch deployed on the second computer device to send a MAC address of the load balancer to the virtual machine that initiates the service, so that the virtual machine that initiates the service sends the load The MAC address of the equalizer is used as the service packet of the destination address.

In a possible design, after the first computer device configures the virtual switch deployed on the second computer device to send the MAC address of the load balancer to the virtual machine that initiates the service, the method further includes:

The virtual machine that initiates the service sends a control packet to the virtual switch, and the control packet Carrying the VIP;

In a possible design, the first computer device creates the load balancing on the second computer device according to the configuration information, including:

In one possible design, the method further includes:

In another aspect, an embodiment of the present invention provides a computer system for implementing load balancing, including at least two computer devices, wherein each computer device includes a central processing unit and a memory, and the at least two computer devices include a first one. a computer device and a second computer device, the first computer device as a cloud management platform, the second computer device running a virtual machine that initiates a service, and the service is a load balancing service, wherein

In a possible design, the forwarding mode of the first computer device configuring the load balancing service on the second computer device includes:

In a possible design, before the virtual machine that sends the service sends the MAC address of the load balancer as the service message of the destination address, the method further includes:

The virtual machine that initiates the service sends the MAC address of the load balancer as the service packet of the destination address, and the virtual switch deployed on the second computer device sends the service according to the destination address of the service packet. The message is sent to the load balancer.

In a possible design, the first computer device deletes the created load balancer after the last virtual machine created on the second computer device that initiates the load balancing service is deleted.

In a possible design, the selected backend server directly returns the result of executing the service to the virtual machine that initiated the service; or

In another aspect, an embodiment of the present invention provides a computer system for implementing load balancing, the computer system including at least three computer devices, wherein each computer device includes a central processing unit and a memory, and the at least three computer devices The first computer device includes a first computer device as a cloud management platform, and the computer device other than the first computer device of the at least three computer devices operates as a business computer device on each business computer device There is a virtual machine that initiates a service, where the service is a load balancing service, where

The first computer device learns that the first virtual machine of the first service computer device is created on the first service computer device, and the load balancer of the first service computer device is created on the first service computer device until Create each of the business calculations on each business computer device Local load balancer on the machine.

In another aspect, an embodiment of the present invention provides a computer system for implementing load balancing, where the computer system includes at least two computer devices, wherein each computer device includes a central processing unit and a memory, and the at least two computer devices The first computer device and the second computer device are included, the first computer device is a cloud management platform, and the second computer device is running a virtual machine that initiates a service, where the service is a load balancing service, where

In a possible design, the first computer device configures the second computer device After the virtual switch deployed on the virtual switch sends the MAC address of the load balancer to the virtual machine that initiates the service, the method further includes:

In one possible design, the method further includes:

To achieve the above objective, a first aspect of the embodiments of the present invention provides a computer device, including a central processing unit CPU, a memory, and a storage device, where the computer device further includes a first virtual machine, a load balancing virtual machine, and a virtual switch. The first virtual machine and the load balancing virtual machine are virtual machines created in the computer device;

The first virtual machine is configured to send a service packet;

The virtual switch is configured to receive the service packet sent by the first virtual machine, and only forward the service packet to the load balancing virtual machine when the service packet needs load balancing processing;

The load balancing virtual machine is configured to receive the service packet forwarded by the virtual switch, and forward the service packet according to a load balancing manner.

After the load balancing virtual machine forwards the service packet based on the load balancing mode, the load balancing virtual machine first determines whether a backend server has been allocated for the service packet after receiving the service packet. Its service, if the backend server has been assigned, then the industry The service message is sent to the backend server that has been assigned. If a backend server has not been assigned, select a backend server to serve it through a load balancing algorithm. The load balancing algorithms include, but are not limited to, polling, weighted polling, random, minimum or source-to-ground hashing, location-based weighted polling, and the like. The location-prioritized weighted polling mode can preferentially select the back-end server in the computer device where the virtual machine that initiates the service packet is located, which can save the speed of the back-end server processing the service packet and save the network bandwidth.

With reference to the first aspect, in a first possible implementation manner of the first aspect of the present disclosure, the that the virtual switch only forwards the service packet to the load balancing virtual machine includes:

The virtual switch sends the MAC address of the load balancing virtual machine to the first virtual machine, so that the first virtual machine sends the service packet based on the obtained MAC address of the load balancing virtual machine. Transmitting, by the virtual switch, the service packet to the load balancing virtual machine; or

The load balancing virtual machine sends its MAC address to the first virtual machine, so that the first virtual machine sends the service packet based on the acquired load balancing virtual machine MAC address, and the virtual switch will Transmitting the service packet to the load balancing virtual machine; or

The first virtual machine sends the service packet based on the pre-configured MAC address of the load balancing virtual machine, and the virtual switch forwards the service packet to the load balancing virtual machine.

In conjunction with the first possible implementation of the first aspect, in a second possible implementation manner of the first aspect of the embodiments, the virtual switch sends the MAC address of the load balancing virtual machine to the first Virtual machines include:

The virtual switch receives the control packet sent by the first virtual machine, and returns a control response packet to the first virtual machine, where the control response packet includes a MAC address of the load balancing virtual machine. That is, the virtual switch can perform the directional processing on the service packet of the first virtual machine according to the configured forwarding mode, so that the first virtual switch forwards the service packet to the first load balancing virtual machine.

With reference to the second possible implementation manner of the first aspect, in a third possible implementation manner of the first aspect of the embodiment, the control packet is an address resolution protocol ARP packet or an Internet control message protocol ICMP packet. . For example, the control packet may be an address resolution protocol ARP packet, and the forwarding mode may be an ARP pickup mode, and the first virtual machine may use the ARP pickup mode to set the MAC address of the first load balancing virtual machine. Sent to the first virtual machine. Optionally, when the control packet is an ICMP packet, the forwarding mode may also be a neighbor discovery protocol NDP pickup mode.

With reference to the first aspect, the first, the second or the third possible implementation manner of the first aspect, in a fourth possible implementation manner of the first aspect of the embodiments, the computer device further includes a virtual load Equalization agent unit;

The virtual load balancing agent unit is configured to receive configuration information for creating a load balancing virtual machine, control creation or deletion of the load balancing virtual machine, and control establishment of a connection between the load balancing virtual machine and the virtual switch. .

And the virtual load balancing agent unit sends the notification information to the virtual machine management unit when the virtual machine management unit is instructed to create the first load balancing virtual machine according to the configuration information of the created load balancing virtual machine, The notification information includes the virtual machine parameter information of the first load balancing virtual machine, where the virtual machine parameter information of the first load balancing virtual machine includes, but is not limited to, the number of virtual central processing units, the size of the virtual memory, and The load balancing information such as a logical network where the virtual machine is located.

The configuration information of the load balancing virtual machine is delivered by the user through the cloud management platform (the second computer device), and the configuration information includes but is not limited to: the identifier of the load balancing virtual machine, and the IP of the load balancing virtual machine. The address, and information about the backend server associated with the load balancing virtual machine. The information of the back-end server includes, but is not limited to, an identifier of the back-end server, an IP protocol (Internet protocol) address of the back-end server, and a method for checking the health of the back-end server. The health check method of the backend server refers to a method in which the load balancing virtual machine checks the health status of the backend server. The back-end server is a server that provides a specific service, and the load balancing virtual machine calculates the service packet according to the received load balance. A load balancing method such as a method selects one of the back-end servers as a server that processes the service message. In the embodiment of the present invention, the backend server may be a virtual machine or a specific physical server.

Optionally, when the first virtual machine is a user-created virtual machine that creates the load balancing virtual machine configuration information, and the first virtual machine is the first user created by the user on the first computer device. The virtual load balancing agent unit starts the creation of the load balancing virtual machine. Similarly, when the last virtual machine configured by the user of the virtual load balancing service is deleted, the first virtual load balancing agent unit starts the deletion of the load balancing virtual machine.

With reference to the first aspect, the first, the second, the third or the fourth possible implementation manner of the first aspect, in a fifth possible implementation manner of the first aspect of the embodiments, a second virtual machine, the second virtual machine having the capability of processing the service message of the first virtual machine;

When the load balancing virtual machine determines that the second virtual machine processes the service packet of the first virtual machine by using a load balancing algorithm, the second virtual machine receives the service forwarded by the load balancing server. Sending a message to the load balancing virtual machine by processing the result of processing the service packet;

The load balancing virtual machine is further configured to send, by the virtual switch, the result of processing the service packet by the second virtual machine to the first virtual machine.

A second aspect of the embodiments of the present invention provides a computer system including a first computer device and a second computer device, wherein the first computer device and the second computer device respectively include a central processing unit CPU, a memory, and a storage device, the first computer device further includes a first virtual machine, a first load balancing virtual machine, and a first virtual switch, where the first virtual machine and the first load balancing virtual machine are in the first a virtual machine created in a computer device;

The first virtual machine is configured to send a service packet;

The first virtual switch is configured to receive the service packet sent by the first virtual machine, When the service packet needs to be load balanced, the service packet is forwarded only to the first load balancing virtual machine;

The first load balancing virtual machine is configured to receive the service packet forwarded by the first virtual switch, and forward the service packet according to a load balancing manner.

After the first load balancing virtual machine forwards the service packet based on the load balancing manner, the first load balancing virtual machine first determines whether a service packet has been allocated for the service packet after receiving the service packet. The backend server serves it, and if a backend server has been assigned, the service message is sent to the already assigned backend server. If a backend server has not been assigned, select a backend server to serve it through a load balancing algorithm. The load balancing algorithms include, but are not limited to, polling, weighted polling, random, minimum or source-to-ground hashing, location-based weighted polling, and the like. The location-prioritized weighted polling mode can preferentially select the back-end server in the computer device where the virtual machine that initiates the service packet is located, which can save the speed of the back-end server processing the service packet and save the network bandwidth.

The second computer device may be a cloud platform management device in the computer system, and implement management and configuration of the first computer device, including configuring and managing a virtual machine in the first computer device.

With reference to the second aspect, in a first possible implementation manner of the second aspect, the first virtual switch that forwards the service packet to the first load balancing virtual machine includes:

Sending, by the first virtual switch, the MAC address of the first load balancing virtual machine to the first virtual machine, so that the first virtual machine sends the MAC address of the first load balancing virtual machine based on the obtained MAC address The first virtual switch forwards the service packet to the first load balancing virtual machine; or

The first load balancing virtual machine sends its MAC address to the first virtual machine, so that the first virtual machine sends the service packet based on the MAC address of the first load balancing virtual machine. Transmitting the service packet to the first load balancing virtual machine by a virtual switch; or

The first virtual machine sends the service packet based on the pre-configured MAC address of the first load balancing virtual machine, and the first virtual switch forwards the service packet to the first load balancing virtual machine.

With reference to the first possible implementation manner of the second aspect, in a second possible implementation manner of the second aspect of the embodiment of the present invention,

The sending, by the first virtual switch, the MAC address of the first load balancing virtual machine to the first virtual machine includes: receiving, by the first virtual switch, a control packet sent by the first virtual machine, to the The first virtual machine returns a control response message, where the control response message includes a MAC address of the first load balancing virtual machine. The forwarding mode of the first virtual switch is configured by the virtual machine management unit, and the first virtual switch performs a directional processing on the service packet of the first virtual machine according to the configured forwarding mode, so that the first virtual The switch forwards the service packet to the first load balancing virtual machine.

With reference to the second possible implementation manner of the second aspect, in a third possible implementation manner of the second aspect of the embodiment, the control packet is an address resolution protocol ARP packet or an Internet control message protocol ICMP packet. . For example, the control packet may be an address resolution protocol ARP packet, and the forwarding mode may be an ARP pickup mode, and the first virtual machine may use the ARP pickup mode to set the MAC address of the first load balancing virtual machine. Sent to the first virtual machine. Optionally, when the control packet is an ICMP packet, the forwarding mode may also be a neighbor discovery protocol NDP pickup mode.

With reference to the second aspect, the first, the second or the third possible implementation manner of the second aspect, in a fourth possible implementation manner of the second aspect of the embodiments, the first computer device further includes a first virtual load balancing agent unit, the second computer device further comprising a virtual load balancing management unit and a virtual machine management unit;

The virtual load balancing management unit is configured to receive configuration information for creating the first load balancing virtual machine, and send the configuration information for creating the first load balancing virtual machine to the first virtual load balancing proxy unit; The first virtual load balancing agent unit is configured to receive the configuration information for creating the first load balancing virtual machine, and instruct the virtual machine management unit to Creating configuration information of the first load balancing virtual machine creates the first load balancing virtual machine.

The first virtual load balancing agent unit sends the notification information to the virtual machine management unit when the virtual machine management unit is instructed to create the first load balancing virtual machine according to the configuration information of the created load balancing virtual machine. The notification information includes the virtual machine parameter information of the first load balancing virtual machine, where the virtual machine parameter information of the first load balancing virtual machine includes but is not limited to: the number of virtual central processing units, and the size of the virtual memory. And information such as the logical network where the load balancing virtual machine is located.

Optionally, the virtual load balancing management unit may directly instruct the virtual machine management unit to create the first load balancing virtual machine according to the configuration information for creating the first load balancing virtual machine.

The configuration information of the first load balancing virtual machine is sent by the user through the cloud management platform (the second computer device), and the configuration information includes but is not limited to: the identifier of the load balancing virtual machine, and the load balancing virtual machine IP address and information about the backend server associated with the load balancing virtual machine. The information of the back-end server includes, but is not limited to, an identifier of the back-end server, an IP protocol (Internet protocol) address of the back-end server, and a method for checking the health of the back-end server. The health check method of the backend server refers to a method in which the load balancing virtual machine checks the health status of the backend server. The back-end server is a server that provides a specific service. The load balancing virtual machine selects one of the back-end servers in a load balancing manner, such as a load balancing algorithm, according to the received service packet, as a server that processes the service packet. In the embodiment of the present invention, the backend server may be a virtual machine or a specific physical server.

With reference to the fourth possible implementation manner of the second aspect, in a fifth possible implementation manner of the second aspect, the first virtual load balancing proxy unit is further configured to indicate the virtual machine management unit And configuring a first virtual network card for the first load balancing virtual machine and establishing a connection between the first virtual network card and the first virtual switch. The virtual machine management unit establishes a connection between the first virtual network card and the first virtual switch. The network identifier of the virtual switch is configured to be implemented, where the network identifier is an identifier of a network interface of the virtual switch. And the virtual machine management unit is configured to create the first virtual network card according to the indication of the first virtual load balancing proxy unit, and further configured to create the first virtual network card according to the configuration information of the load balancing virtual machine. The virtual network card assigns an IP address and a MAC address.

In conjunction with the fourth or fifth possible implementation of the second aspect, in a sixth possible implementation manner of the second aspect of the embodiments, the first virtual machine is configured to create the first load balancing virtual machine a virtual machine created by the user of the configuration information; the virtual load balancing management unit is further configured to receive configuration information for creating the first virtual machine, and determine that the first virtual machine is the user at the first Notifying the first virtual load balancing agent unit of the information of the first virtual machine when the first virtual machine is created on the computer device; the first virtual load balancing agent unit is configured to receive the When the notification of the first virtual machine is sent by the virtual load balancing management unit, the virtual machine management unit is instructed to create the first load balancing virtual machine. The virtual load balancing management unit may also directly notify the virtual machine management unit to create the first load balancing virtual machine.

Similarly, when the last virtual machine configured by the user of the virtual load balancing service is deleted, the virtual load balancing management unit notifies the first virtual load balancing agent unit to instruct the virtual machine management unit to delete the The first load balancing virtual machine, or directly notifying the virtual machine management unit to delete the first load balancing virtual machine.

With reference to the second aspect, the first to the sixth possible implementation manner of the second aspect, in the seventh possible implementation manner of the second aspect of the embodiments, the first computer device further includes a second a virtual machine, the second virtual machine having the capability of processing the service packet of the first virtual machine;

When the first load balancing virtual machine determines that the second virtual machine processes the service packet of the first virtual machine by using a load balancing manner, the second virtual machine receives the first load balancing virtual machine. Forwarding the service packet, and sending the result of processing the service packet to the first load balancing virtual machine;

The first load balancing virtual machine is further configured to send, by the second virtual machine, the result of processing the service packet to the first virtual machine by using the first virtual switch.

With reference to the second aspect, the first to the seventh possible implementation manners of the second aspect, in an eighth possible implementation manner of the second aspect of the embodiments, the computer system further includes a third computer device The third computer device includes a second virtual load balancing agent unit, a second virtual switch, and a third virtual machine;

The second virtual load balancing proxy unit is configured to receive configuration information that is sent by the virtual load balancing management unit to create a second load balancing virtual machine, and to indicate the configuration information based on the configuration information of the second load balancing virtual machine. The virtual machine management unit creates the second load balancing virtual machine, and instructs the virtual machine management unit to configure a second virtual network card for the second load balancing virtual machine and establish the second virtual network card and the second virtual The connection between the switches;

The second virtual switch is configured to receive the second service packet initiated by the third virtual machine, and when the second service packet needs load balancing processing, only forward the second service packet to the a second load balancing virtual machine;

The second load balancing virtual machine is configured to receive the second service packet that is forwarded by the second virtual switch, and forward the second service packet according to a load balancing manner.

In conjunction with the eighth possible implementation of the second aspect, in a ninth possible implementation manner of the second aspect of the embodiments, the virtual machine management unit configures the first IP address of the first load balancing virtual machine And configuring a second IP address and a second MAC address for the second load balancing virtual machine; wherein the first IP address is the same as the second IP address, or the first MAC address The address is the same as the second MAC address.

Optionally, the first computer device may include hardware devices such as network hardware in addition to the CPU, memory, and storage device; the network hardware may be a switching device (such as a network card), a router, and/or other network. A device is used to implement communication between multiple devices. Multiple devices can be connected by wireless or wired connection.

With reference to the second aspect, any one of the first to the ninth possible implementation manners of the second aspect, In a tenth possible implementation manner of the second aspect, the second computer device further includes a virtual load balancing storage device, configured to store configuration information for creating a load balancing virtual machine. The virtual load balancing storage device may be a storage device in the second computer device, or may be a storage device independent of the second computer device.

A third aspect of the embodiments of the present invention provides a computer system, including a first computer device, a second computer device, and a third computer device, wherein the first computer device, the second computer device, and the third computer device respectively The central computer includes a CPU, a memory, and a storage device, the second computer device configuring and managing a virtual machine on the first computer device and the third computer device, the first computer device including a first virtual a switch, the third computer device includes a second virtual switch, the second computer device configured to receive configuration information for creating a load balancing virtual machine on the first computer device, created in the first computer device First, load balancing the virtual machine, and establishing a connection between the first load balancing virtual machine and the first virtual switch; and receiving configuration information for creating a load balancing virtual machine on the third computer device, where Creating a second load balancing virtual machine in the three computer devices, and establishing the second load balancing virtual machine The virtual switch connected to the second;

The first virtual switch is configured to receive a first service packet sent by the first virtual machine in the first computer device, and only forward the first service packet when the first service packet needs load balancing processing Transmitting a service packet to the first load balancing virtual machine;

The second virtual switch is configured to receive a second service packet sent by the second virtual machine in the third computer device, and only forward the second service packet when the second service packet needs load balancing processing The service packet is sent to the second load balancing virtual machine.

With reference to the third aspect, in a first possible implementation manner of the third aspect, the first virtual switch that forwards the first service packet to the first load balancing virtual machine includes:

Sending, by the first virtual switch, the MAC address of the first load balancing virtual machine to the first virtual machine, so that the first virtual machine is based on the acquired first load balancing virtual The MAC address of the machine sends the service packet, and the first virtual switch forwards the service packet to the first load balancing virtual machine; or

With reference to the first possible implementation manner of the third aspect, in a second possible implementation manner of the third aspect, the first virtual switch sends the MAC address of the first load balancing virtual machine to The first virtual machine includes:

Receiving, by the first virtual switch, a control packet sent by the first virtual machine, and returning a control response packet to the first virtual machine, where the control response packet includes a MAC address of the first load balancing virtual machine. . The forwarding mode of the first virtual switch is configured by the virtual machine management unit, and the first virtual switch performs a directional processing on the service packet of the first virtual machine according to the configured forwarding mode, so that the first virtual The switch forwards the service packet to the first load balancing virtual machine.

With reference to the second possible implementation manner of the third aspect, in a third possible implementation manner of the third aspect of the embodiments, the control packet is an address resolution protocol ARP packet or an Internet control message protocol ICMP packet. . For example, the control packet may be an address resolution protocol ARP packet, and the forwarding mode may be an ARP pickup mode, and the first virtual machine may use the ARP pickup mode to set the MAC address of the first load balancing virtual machine. Sent to the first virtual machine. Optionally, when the control packet is an ICMP packet, the forwarding mode may also be a neighbor discovery protocol NDP pickup mode.

With reference to the third aspect, the first to the third possible implementation manners of the third aspect, the fourth possible implementation manner of the third aspect of the embodiment of the present invention, the first computer device further includes a second a virtual machine, the second virtual machine having the service report for processing the first virtual machine Ability of the text;

When the first load balancing virtual machine determines that the second virtual machine processes the service packet of the first virtual machine by using a load balancing manner, the second virtual machine receives the first load balancing server for forwarding. The service packet, and the result of processing the service packet is sent to the first load balancing virtual machine;

With reference to the third aspect, any one of the first to fourth possible implementation manners of the third aspect,

The IP address of the first load balancing virtual machine is the same as the IP address of the second load balancing virtual machine; or

The MAC address of the first load balancing virtual machine is the same as the MAC address of the second load balancing virtual machine.

In a fourth aspect, an embodiment of the present invention provides a method for implementing load balancing, including:

The second computer device receives configuration information for creating the load balancing virtual machine on the first computer device, the first computer device including a central processing unit CPU, a memory, and a storage device, the first computer device further including a first virtual Machine and virtual switch;

The second computer device creates a load balancing virtual machine on the first computer device according to the configuration information, and establishes a connection between the load balancing virtual machine and a virtual switch in the first computer device;

The virtual switch receives the service packet sent by the first virtual machine, and only forwards the service packet to the load balancing virtual machine when the service packet needs load balancing processing;

The load balancing virtual machine receives the service packet forwarded by the first virtual switch, and forwards the service packet according to a load balancing manner.

The load balancing virtual machine forwards the service packet based on the load balancing mode. The load balancing virtual machine first determines whether the service packet is the service packet after receiving the service packet. A backend server is assigned to serve it, and if a backend server has been assigned, the service message is sent to the already assigned backend server. If a backend server has not been assigned, select a backend server to serve it through a load balancing algorithm. The load balancing algorithms include, but are not limited to, polling, weighted polling, random, minimum or source-to-ground hashing, location-based weighted polling, and the like. The location-prioritized weighted polling mode can preferentially select the back-end server in the computer device where the virtual machine that initiates the service packet is located, which can save the speed of the back-end server processing the service packet and save the network bandwidth.

With reference to the fourth aspect, in a first possible implementation manner of the fourth aspect of the embodiments of the present disclosure, the that the virtual switch only forwards the service packet to the load balancing virtual machine includes:

The load balancing virtual machine sends its MAC address to the first virtual machine, so that the first virtual machine sends the service packet based on the obtained MAC address of the load balancing virtual machine, where the virtual switch Forwarding the service packet to the load balancing virtual machine; or

With reference to the first possible implementation manner of the fourth aspect, in a second possible implementation manner of the fourth aspect, the virtual switch sends the MAC address of the load balancing virtual machine to the first Virtual machines include:

The virtual switch receives the control packet sent by the first virtual machine, and returns a control response packet to the first virtual machine, where the control response packet includes a MAC address of the load balancing virtual machine. The forwarding mode of the virtual switch is configured, and the virtual switch performs directional processing on the service packet of the first virtual machine according to the configured forwarding mode, so that the first The virtual switch forwards the service packet to the load balancing virtual machine.

In conjunction with the second possible implementation manner of the fourth aspect, in a third possible implementation manner of the fourth aspect, the control packet is an address resolution protocol ARP packet or an Internet control message protocol ICMP packet. . For example, the control packet may be an address resolution protocol ARP packet, and the forwarding mode may be an ARP pickup mode, and the first virtual machine may use the ARP pickup mode to set the MAC address of the first load balancing virtual machine. Sent to the first virtual machine. Optionally, when the control packet is an ICMP packet, the forwarding mode may also be a neighbor discovery protocol NDP pickup mode.

With reference to the fourth aspect, any one of the first to the third possible implementation manners of the fourth aspect, the fourth possible implementation manner of the fourth aspect of the embodiments of the present invention, The method further includes: before the creating a load balancing virtual machine on the first computer device, the method further includes:

Sending, by the second computer device, the configuration information to the first computer device;

The first computer device receives configuration information for creating a load balancing virtual machine, instructing the second computer device to create the load balancing virtual machine. Optionally, the second computer device may directly create the load balancing virtual machine on the second computer device according to the configuration information.

The configuration information of the load balancing virtual machine is delivered by the user through the cloud management platform (the second computer device), and the configuration information includes but is not limited to: the identifier of the load balancing virtual machine, and the IP of the load balancing virtual machine. The address, and information about the backend server associated with the load balancing virtual machine. The information of the back-end server includes, but is not limited to, an identifier of the back-end server, an IP protocol (Internet protocol) address of the back-end server, and a method for checking the health of the back-end server. The health check method of the backend server refers to a method in which the load balancing virtual machine checks the health status of the backend server. The back-end server is a server that provides a specific service. The load balancing virtual machine selects one of the back-end servers in a load balancing manner, such as a load balancing algorithm, according to the received service packet, as a server that processes the service packet. In the embodiment of the present invention, the backend server may be a virtual machine or a Physical server.

With reference to the fourth aspect, any one of the first to fourth possible implementation manners of the fourth aspect, in the fifth possible implementation manner of the fourth aspect of the embodiments, the second computer device further includes the second a virtual machine, the second virtual machine is capable of processing the service packet of the first virtual machine, and the method further includes:

The load balancing virtual machine sends the result of processing the service packet by the second virtual machine to the first virtual machine by using the virtual switch.

The computer device, the system, and the load balancing method provided by the embodiments of the present invention, when a load balancing virtual machine is created in each computer device, when the service of the virtual machine in the computer device needs load balancing processing, the related service packet is only A load balancing virtual machine that is sent to the computer device performs load balancing processing. Since the load balancing related service packets are implemented in the computer device, the service packets on the plurality of computer devices are not concentrated to one load balancing node, thereby avoiding the processing delay caused by the centralized load balancing congestion. Moreover, when the computer device fails, the virtual machine in the computer device will also go offline, and there will be no need for load balancing. Therefore, the load balancing virtual machine failure caused by the failure of a single computer device does not affect the load balancing virtual machine on other computer devices, and there is no service interruption caused by the load balancing virtual machine failure. .

DRAWINGS

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below. Obviously, the drawings in the following description are only It is some embodiments of the present invention, and those skilled in the art can obtain other ones according to these drawings without paying creative labor. The figure.

1 is a schematic diagram of implementing a load balancing architecture in the prior art;

2 is a schematic structural diagram of implementing an implementation of virtual load balancing according to an embodiment of the present invention;

3 is a schematic structural diagram of a computer device according to an embodiment of the present invention;

4 is a structural diagram of a specific implementation structure of a computer device according to an embodiment of the present invention;

FIG. 5 is a schematic structural diagram of an implementation of a first computer system according to an embodiment of the present invention; FIG.

6 is a structural diagram of a specific implementation structure of a computer system in a first embodiment of the present invention;

FIG. 7 is a structural diagram showing another specific implementation structure of the first computer system according to an embodiment of the present invention; FIG.

FIG. 8 is a schematic flowchart of a method for implementing load balancing according to an embodiment of the present invention;

FIG. 9 is a schematic flowchart of another method for implementing load balancing according to an embodiment of the present invention.

detailed description

The technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present invention. It is obvious that the described embodiments are a part of the embodiments of the present invention, but not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of the present invention.

Referring to FIG. 1, FIG. 1 is a schematic diagram of implementing a load balancing architecture in the prior art. 1 includes a computer device 101, a computer device 102, a computer device 103, a primary virtual load balancer 104, and a standby virtual load balancer 105. The computer device 101 functions as a cloud management platform, including a central processing unit (CPU) 1011 and a storage device 1012 (such as a hard disk, a solid state hard disk, etc.), and runs a cloud management platform related program to implement the pair of computer devices 102. Management and configuration of nodes such as computer device 103. The computer device 102 and the computer device 103 are connected by a network, and the computer device 101 can create a corresponding virtual machine at the computer device 102 or the computer device 103 according to the received configuration information about creating a virtual machine.

FIG. 1 illustrates an example in which each computer device includes two virtual machines. For example, computer device 102 includes virtual machines 10211 and 10212, and computer device 103 includes virtual machines 10311 and 10312. The primary load balancer 104 and the backup load balancer 105 provide load balancing services for service messages sent by the virtual machines in the computer device 102 or the computer device 103 in a master-slave manner. The primary load balancer 104 and the backup load balancer 105 are connected by heartbeat detection. When the primary load balancer 104 fails, the backup load balancer 105 takes over the primary load balancer 104 to continue to provide load balancing services for the virtual machine's request.

In the load balancing mode, when the primary load balancing node 104 needs to be switched to the standby load balancing node, the load balancing service is not provided by the load balancing device, and the service packet that needs to be load balanced cannot be processed. At the same time, only one load balancer (load balancer 104 or 105) processes the service packets of all the virtual machines, which is easy to cause overload and congestion of the virtual load balancer, and the timely processing of the affected request services reduces the user experience.

The method, device, and system for virtual load balancing are provided by creating a distribution, in order to solve the problem that the virtual load balancing uses the interruption caused by the active/standby mode during the handover, and the congestion caused when the virtual load balancing node is centralized. The load balancing function is used to offload the load balancing function to the device side of each node. The centralized load balancing mode of the active/standby switchover is not required to avoid the interruption caused by the active/standby switchover and the congestion caused by centralized load balancing.

In order to clearly describe the technical solution, the embodiment of the present invention uses a cloud management platform and a network composed of two computer device nodes as an example for description. As shown in FIG. 2, FIG. 2 is a schematic diagram of an implementation architecture for implementing virtual load balancing according to an embodiment of the present invention. The computer device 201 functions as a cloud management platform to implement management and configuration of the virtualization of the computer device 202 and the computer device 203. It should be noted that, in FIG. 2, two computer devices are taken as an example to describe the solution of the embodiment of the present invention. In a specific implementation, the computer device 201 as a cloud platform can also configure and manage two or more computer devices. The example does not limit the number of specific computer devices.

Taking the computer device 202 in FIG. 2 as an example, a group of computer devices in the embodiment of the present invention Explain in part. The computer device 202 can include a hardware resource layer 2023 comprised of computing hardware 20231, storage hardware 20232, and network hardware 20233. Computing hardware 20231 can be a dedicated processor or a general purpose processor for providing processing and computing functions. The storage hardware 20232 is configured to provide storage capabilities, which may be provided by the storage hardware 20232 itself (for example, a server's local memory or a local hard disk), or may be provided through a network (for example, the server connects to a network storage device through a network). The network hardware 20233 may be a switching device (network card, etc.), a router, and/or other network device. The network hardware 20233 is used to implement communication between a plurality of devices, and the plurality of devices are connected by wireless or wired. The virtualization layer in computer device 202 is used to abstract the hardware resources of the hardware resource layer to provide virtual resources. The operating system 2022 in the computer device 202 includes, but is not limited to, a virtualization layer Hypervisor 20221 and a virtual switch 20222 running virtualization management. The virtualization layer Hypervisor 20221 manages virtual machines (e.g., virtual machines 20213, etc.) in the computer device 202. The virtual switch 20222 provides a network connection function between the virtual machines managed by the Hypervisor 20221 and between the virtual machine and the physical network. The virtual load balancing agent unit 20211 connects the load balancing virtual machine 20212 in the computer device 202 through the virtualization layer Hypervisor 20221. Configure and manage.

The computer device 201 functions as a cloud management platform, runs a cloud management platform related program, implements management and configuration of the computer device 202 and the computer device 203, and includes configuring and managing virtual machines in the computer device 202 and the computer device 203. . The computer device 202 and the computer device 203 serve as device nodes under the control of the cloud management platform, and provide related services or services such as virtual machines.

It should be noted that, in FIG. 2, the computer device 201 includes a hardware resource layer 2013 composed of a computing hardware 20131, a storage hardware 20132, and a network hardware 20133. The computing hardware 20131 can be a dedicated processor or a general purpose processor for providing processing and computing functions. The storage hardware 20132 is used to provide storage capacity, which may be provided by the storage hardware 20132 itself (for example, a server's local memory or a local hard disk), or may be provided through a network (for example, a service) Connect a network storage device through the network). The network hardware 20133 may be a switching device (network card, etc.), a router, and/or other network devices. The network hardware 20133 is used to implement communication between multiple devices, and multiple devices are connected by wireless or wired. The virtualization layer in computer device 201 is used to abstract the hardware resources of the hardware resource layer to provide virtual resources. The operating system 2012 in the computer device 201 includes, but is not limited to, a virtualization layer Hypervisor 20121 running virtualization management.

The following describes the implementation of the load balancing provided by the embodiment of the present invention by using the cloud management platform to apply for a virtual machine on the computer device 202 and load balancing the service packets of the applied virtual machine.

In the embodiment of the present invention, the load balancer can perform load balancing processing on the service packets of the applied virtual machine, that is, the load balancer is configured to perform load balancing processing on the service packets of the applied virtual machine. The load balancer includes, but is not limited to, a virtual machine, a namespace, or a container. The embodiment of the present invention uses a virtual machine as a load balancer as an example for description. It can be understood that the implementation of using a namespace or a container as a load balancer can be implemented by referring to an implementation in which a virtual machine is used as a load balancer.

In the embodiment of the present invention, the cloud management platform implements unified scheduling management of resources such as network, storage, and computing, and the business automation orchestration on the above, realizes the operation functions of charging, process, log, and application delivery, and finally presses A pay-as-you-go model that provides users, administrators with available, convenient, on-demand network access, giving users and administrators access to a configurable pool of computing resources (resources including network, server, storage, application software, Service, etc.).

Referring to FIG. 2, the computer device 201 as a cloud management platform includes a virtual load balancing management unit 20111 and a virtual machine management unit 20112.

The virtual load balancing management unit 20111 is configured to receive configuration information for creating a load balancing virtual machine, and store the configuration information of the created load balancing virtual machine into the virtual load balancing data device 20114. The virtual load balancing management unit 20111 may further send the configuration information of the created load balancing virtual machine to the virtual load balancing proxy unit 20211.

The virtual machine management unit 20112 is configured to create or delete the load balancing virtual machine 20212 according to the configuration information of the created load balancing virtual machine.

The configuration information of the load balancing virtual machine is delivered by the user through the cloud management platform, and the configuration information includes, but is not limited to, an identifier of the load balancing virtual machine, and an IP address of the load balancing virtual machine, such as a virtualized IP address. (VIP, Virutal IP Address) address, and information about the backend server associated with the load balancing virtual machine. The VIP address of the load balancing virtual machine is an IP address of the load balancing virtual machine. The information of the back-end server includes, but is not limited to, an identifier of the back-end server, an IP protocol (Internet protocol) address of the back-end server, and a method for checking the health of the back-end server. The health check method of the backend server refers to a method in which the load balancing virtual machine checks the health status of the backend server. The back-end server is a server that provides a specific service. The load balancing virtual machine selects one of the back-end servers in a load balancing manner, such as a load balancing algorithm, according to the received service packet, as a server that processes the service packet. In the embodiment of the present invention, the backend server may be a virtual machine or a specific physical server.

In FIG. 2, the virtual load balancing data device 20114 is configured to store configuration information for creating a load balancing virtual machine, which may be a storage device in the computer device 201, or may be a storage device independent of the computer device 201, and the present invention The embodiment does not limit the specific manner of saving the configuration information of the created load balancing virtual machine.

The virtual machine management unit 20112 may directly create or delete the load balancing virtual machine 20212 according to the instruction of the virtual load balancing management unit 20111; or may create or delete according to the instruction of the virtual load balancing agent unit 20211 in the computer device 202. The load balancing virtual machine 20212.

The virtual machine management unit 20112 is further configured to create a corresponding virtual machine according to the configuration information that is created by the user and create a virtual machine in the computer device 202. For example, when the computer device 201 receives a virtual machine created by the user in the computer device 202, the virtual machine management unit 20112 creates a virtual machine 20213 in the computer device 202. Similarly, when the meter When the computer device 201 receives the configuration information of the virtual machine created by the user as the backend server in the computer device 202, the virtual machine management unit 20112 creates a corresponding virtual machine as a backend server, for example, creates a virtual machine 20214. .

As an optional implementation, the computer device 202 may further include a virtual load balancing proxy unit 20211, where the virtual load balancing proxy unit 20211 is configured to obtain the created load balancing virtuality from the virtual load balancing management unit 20111. The configuration information of the machine instructs the virtual machine management unit 20112 to create a load balancing virtual machine 20212 according to the configuration information of the created load balancing virtual machine.

The virtual load balancing agent unit 20211, when the virtual machine management unit 20112 is instructed to create the load balancing virtual machine 20212 according to the configuration information of the created load balancing virtual machine, sends the load balancing virtual machine 20212 to the virtual machine management unit 20112. Notification information, the notification information further including virtual machine parameter information for creating a load balancing virtual machine 20212. The virtual machine parameter information of the load balancing virtual machine 20212 includes, but is not limited to, the number of virtual central processing units, the size of the virtual memory, and the logical network in which the load balancing virtual machine 20212 is located.

The process of creating the load balancing virtual machine 20212 by the virtual machine management unit 20112 may be: the virtual machine management unit 20112 specifies an image with a load balancing module by calling an application programming interface (API) of the created virtual machine. A load balancing virtual machine 20212 is created. Virtual machine parameter information is used when calling the API to create a virtual machine to create a virtual machine based on virtual machine parameter information. The mirror with the load balancing module is specified, so that the created virtual machine has a load balancing module, and then the configuration information is configured to the load balancing module in the virtual machine.

When the virtual load balancing management unit 20111 directly notifies the virtual machine management unit 20112 to create the load balancing virtual machine 20212, the virtual load balancing management unit 20111 sends a notification message to the virtual machine management unit 20112, where the notification message includes creating the The configuration information and virtual machine parameter information required by the virtual machine 20212 are load balanced to notify the virtual machine management unit 20112 to start creating the load balancing virtual machine 20212. Already created in the computer device 202 After the load balancing virtual machine 20212, the load balancing virtual machine 20212 is configured to receive a virtual machine in the computer device 202, for example, the virtual machine 20213, and the service packet sent by the virtual switch 20222 is received based on the load balancing manner. The business message is processed.

The load balancing virtual machine 20212 processes the service packet of the virtual machine in the computer device 202 based on the load balancing mode. The load balancing virtual machine 20212 first determines whether it is already after receiving the service packet. The service packet is assigned a backend server for its service. If the backend server has been allocated, the service packet is sent to the already assigned backend server. If a backend server has not been assigned, select a backend server to serve it through a load balancing algorithm. The load balancing algorithms include, but are not limited to, polling, weighted polling, random, minimum or source-to-ground hashing, location-based weighted polling, and the like. The location-prioritized weighted polling mode can preferentially select the back-end server in the computer device where the virtual machine that initiates the service packet is located, which can save the speed of the back-end server processing the service packet and save the network bandwidth.

In the embodiment of the present invention, a polling algorithm is taken as an example, and a manner in which the load balancing virtual machine 20212 selects a backend server is described. When the backend server of the load balancing virtual machine 20212 includes the virtual machine virtual machine 20214 and the virtual machine 20314, the load balancing virtual machine 20212 groups all the backend servers (the virtual machine 20214 and the virtual machine 20314) into a list, taking turns in the list. Allocate business messages that need to be processed. When the load balancing virtual machine 20212 receives the first service packet sent by the virtual machine 20213, the virtual machine 20214 is used as the backend server for providing the service, and sends the first service packet to the virtual machine 20214; When the second service message from the virtual machine 20213 is received, the virtual machine 20314 is used as the back-end server for providing the service, and then the second service message is sent to the virtual machine 20314; when the load balancing virtual machine 20212 receives The third service packet sent by the virtual machine 20213 is used as the backend server for providing the service, and the third service packet is sent to the virtual machine 20214 for circulation.

In the above-mentioned load balancing mode, the user applies for a virtual machine for initiating a service in the computer device 202 through the cloud management platform, and needs to load the service request of the applied virtual machine. The equalization process is described as an example. It can be understood that when a user applies for a virtual machine for initiating a service in the computer device 203 through the cloud management platform, and needs to perform load balancing processing on the service request of the applied virtual machine, the implementation manner is similar. Since the load balancing unit is distributed in each computer device, there is no centralized load balancing node, and the service packets related to load balancing are load-balanced in each computer device, and there is no congestion caused by service packets. Processing delay. Moreover, a load balancer is deployed in both the computer device 202 and the computer device 203. When a certain computer device fails, the virtual machine in the computing system is also offline, and no load balancing requirement is generated; The failure of the load balancing unit caused by the failure of a single computer device does not affect the processing of other load balancing services, and there is no service interruption caused by the active/standby switchover caused by the failure of the load balancing node.

In the embodiment of the present invention, the configuration information of the load balancing virtual machine 20212 obtained by the virtual load balancing proxy unit 2011 from the virtual load balancing management unit 20111 may further include network information of the load balancing virtual machine 20212. The virtual load balancing agent unit 20211 is further configured to instruct the virtual machine management unit 20112 to establish a virtual network card for the load balancing virtual machine 20212 according to the network information in the configuration information of the load balancing virtual machine, and indicate the virtual The machine management unit 20112 establishes a connection between the virtual network card and the virtual switch 20222. The virtual machine management unit 20112 may also allocate an IP address to the created virtual network card according to the configuration information of the load balancing virtual machine when the virtual network card is established according to the instruction of the virtual load balancing agent unit 20211. MAC address. Optionally, the virtual machine management unit 20111 may be instructed by the virtual load balancing management unit 20111 to establish a virtual network card for the load balancing virtual machine 20212, and establish a connection between the virtual network card and the virtual switch 20222.

Establishing a connection between the virtual network card and the virtual switch 20222 can be implemented by configuring a network identifier of the virtual switch. The network identifier of the virtual switch is an identifier of the network interface of the virtual switch. When the network interface of the virtual machine is the same as the identifier of the network interface of the virtual switch, the virtual machine and the virtual switch can establish a connection. In the implementation of the present invention, the virtual The machine management unit 20112 establishes a connection between the load balancing virtual machine 20212 and the virtual switch 20222 by configuring the network identifier of the virtual switch 20222. For example, the virtual switch 20222 has three network interfaces, and the network identifier is 100. When the virtual network card of the load balancing virtual machine 20212 and the virtual machine 20213 has an interface with the network identifier of 100, the virtual machine 20213 can be identified by the network as an interface of 100. The communication with the interface of the load balancing virtual machine 20212 having the network identifier of 100 is implemented. The virtual machine management unit 20112 may configure a network identifier for the virtual switch 20222 according to the virtual load balancing proxy unit 20211 or the virtual load balancing management unit 20111. It can be understood that the three network interfaces of the virtual switch 20222 are only examples. In a specific implementation, the virtual switch 20222 may further include other network interfaces. For example, the virtual switch 20222 includes 100 network interfaces, and the network identifier of the three network interfaces is 100.

After the virtual machine management unit 20112 establishes the connection between the virtual network card and the virtual switch 20222 according to the instruction of the virtual load balancing agent unit 20211, the virtual load balancing agent unit 20211 is further configured to instruct the virtual machine management unit. The configuration of the virtual switch 20222 is configured to forward the service packet of the virtual machine 20213 according to the forwarding mode, so that the virtual machine 20213 sends the service packet to the The load balancing virtual machine 20212. Optionally, the virtual machine management unit 20112 may be instructed by the virtual load balancing management unit 20111 to configure a forwarding mode of the virtual switch 20222.

The forwarding mode of the virtual switch 20222 may be an ARP address resolution protocol (ARP) or a NDP Neighbor Discovery Protocol (NDP) mode, that is, the virtual switch 20222 is configured to the service packet. The way to perform directional forwarding processing is ARP or NDP pickup mode. The embodiment of the present invention uses a cloud platform supporting OpenStack as an example to describe the implementation manner of the ARP pickup technology.

The ARP pickup technology is implemented by a flow table. The flow table technology generally consists of two parts, one is the matching condition and the other is the action. The matching condition of the flow table of the ARP pickup is: the protocol type is ARP, the destination protocol address of the ARP is the virtualized IP address (VIP, Virutal IP Address), which is the IP address that the load balancing virtual machine can access, and the ARP operation type is the ARP request. The action includes: 1. Modifying the destination media access control (MAC, Media Access Control) address of the received ARP packet as the source MAC address of the packet; 2. Modifying the source MAC address of the received ARP packet as load balancing. The MAC address of the virtual machine 20212; 3. The ARP operation type of the modified ARP packet is an ARP response; 4. The ARP destination protocol address of the received ARP packet is the ARP source protocol address of the ARP packet, where The protocol address is the term in the ARP protocol, that is, the IP address. 5. The ARP source protocol address of the received ARP packet is the IP address of the load balancing virtual machine 20212. 6. The ARP destination of the received ARP packet is modified. The hardware address is the ARP source hardware address of the packet. The hardware address is the term in the ARP protocol, that is, the MAC address. 7. The ARP source hardware address of the received ARP packet is the MAC address of the load balancing virtual machine 20212. . The actions in the above flow table are not necessarily performed for each action, and those skilled in the art will recognize that action 2 is necessary. Then, the modified ARP packet is sent to the virtual machine 20213 by the virtual switch 20222. The ARP proxy mode of the virtual switch 20222 enables the virtual machine 20213 to obtain the MAC address corresponding to the VIP, and sends a service packet based on the obtained MAC address, so that the virtual machine in the computer device 202 needs to perform load balancing processing. The message is only sent to the load balancing virtual machine 20212 on the computer device 202.

After the virtual switch 20222 sends the ARP reply packet to the virtual machine 20213, the virtual machine 20213 constructs a service request packet. The IP address of the service request packet is a VIP address, and the MAC address is the load obtained from the ARP response of the virtual switch 20222. The MAC address of the virtual machine 20212 is equalized. The virtual machine 20213 sends the configured service request message to the virtual switch 20222. The virtual switch 20222 sends the service request message to the load balancing virtual machine 20212 according to the IP address and the MAC address of the service request message.

For the packet sent by the virtual machine 20213 that does not require load balancing, the IP address is not a VIP address, and the virtual switch 20222 can forward according to the normal forwarding mode.

The virtual switch 20222 needs to enter the service packet initiated by the virtual machine 20213. During the line load balancing process, the virtual switch 20222 sends the service message of the virtual machine 20213 to the load balancing virtual machine 20212 and the load balancing virtual machine on other computer devices (such as the computer device 203) through the ARP pickup technology ( If the load balancing virtual machine 20312 does not receive the ARP request, the service packet sent by the virtual machine 2013 in the computer device 202 is not received, and the service packet of the virtual machine in each computer device needs load balancing. When processed, it is only implemented on a load balancing virtual machine within the computer device.

Similar to creating a load balancing virtual machine 20212 in the computer device 202, a load balancing virtual machine 20312 can also be created in the computer device 203. Since the service packets sent by the virtual machine in each computer device that require load balancing processing are forwarded only to the load balancing virtual machine created in the computer device, they are not forwarded to the load balancing virtual machine in other computer devices for processing. The IP address of the load balancing virtual machine 20212 in the computer device 202 may be the same as the IP address of the load balancing virtual machine 20312 in the computer device 203. Similarly, the MAC address of the load balancing virtual machine 20212 in the computer device 202 may be associated with the computer device. The MAC address of the load balancing virtual machine 20312 is the same in 203. In this way, IP address resources and/or MAC address resources in the network formed by the computer devices can be saved.

In the embodiment of the present invention, the virtual load balancing proxy unit 20211 may instruct the virtual machine management unit 20112 to create a load balancing virtual machine when the user deploying the load balancing service creates the first virtual machine in the computer device 202. 20212. This can save resources of the computer device 202 and avoid waste of hardware resources caused by creating a load balancing virtual machine when the load balancing service is not required. The timing of creating the first virtual machine in the computer device 202 by the user deploying the load balancing service may be the virtual load balancing management when the virtual machine management unit 20112 creates a virtual machine (for example, the virtual machine 20213). The unit 20111 determines whether the created virtual machine is the first virtual machine configured by the user who has configured the virtual load balancing service. If so, the virtual load balancing management unit 20111 notifies the virtual load balancing agent unit to instruct the virtual machine management unit 20112 to create the load balancing virtual machine 20212. The virtual load balancing management unit 20111 may directly notify the virtual machine management unit 20112. A load balancing virtual machine 20212 is created. When the last virtual machine configured by the user of the virtual load balancing service is deleted, the virtual load balancing management unit 20111 notifies the virtual load balancing agent unit to instruct the virtual machine management unit 20112 to delete the load balancing virtual The machine 20212, or directly notifying the virtual machine management unit 20112, deletes the load balancing virtual machine 20212.

The load balancing virtual machine 20212 is created when the user who has configured the virtual load balancing service creates the first virtual machine, and the load balancing virtual machine 20212 is deleted when the last virtual machine configured by the same user is deleted, thereby saving computer equipment. The virtual resources in 202 increase the utilization of resources in computer device 202.

The creation and configuration of the load balancing virtual machine 20212 described above is that the virtual load balancing agent unit 20211 in the computer device 202 instructs the virtual machine management unit 20112 in the computer device 201 to create. In a specific implementation, the virtual load balancing agent unit 20211 may not be included in the computer device 202. When the virtual load balancing management unit 20111 in the computer device 201 receives the configuration information for creating the load balancing virtual machine 20212, The virtual machine management unit 20112 is directly instructed to create a load balancing virtual machine 20212. The embodiment of the present invention does not limit the manner in which the specific load balancing virtual machine 20212 is created.

A user applies for the virtual machine 20213 through the cloud management platform, and accesses the video service through the virtual machine 20213. The video service can be provided by the back end server such as the virtual machine 20214 or the virtual machine 20314 as an example to illustrate the implementation of load balancing in the embodiment of the present invention. . In conjunction with the embodiment shown in FIG. 2, the cloud management platform (computer device 201) creates a load balancing virtual machine 20212 in the computer device 202, and creates a load balancing virtual machine 20312 in the computer device 203; when the virtual machine 20213 initiates access to the video service, The service message is forwarded only by the virtual switch 20222 to the load balancing virtual machine 20212 in the computer device 202, and the load balancing virtual machine 20312 in the computer device 203 connected to the computer device 202 through the network does not receive the service message. The service packet sent by the virtual machine 20213 causes the service packet initiated by the computer device 202 to be load balanced to perform load balancing processing only in the computer device 202. Load balancing virtual machine 20212 After receiving the service packet of the virtual machine 20213 forwarded by the virtual switch 20222 for the video service, the service packet is processed by the load balancing algorithm. In the backend server, the virtual machine 20214 or the virtual machine 20314 may be selected as the virtual machine 20213. Provide video services. Preferably, the load balancing virtual machine 20212 can select the virtual machine 20214 as a backend server to provide a video service to the virtual machine 20213. This can save network bandwidth and improve the efficiency of the backend server in responding to service messages.

Referring to FIG. 3, FIG. 3 is a schematic structural diagram of a computer device 300 according to an embodiment of the present invention. In FIG. 3, the computer device 300 includes a central processing unit CPU 3031, a memory 3032, and a storage device 3033, and further includes a first virtual machine 304, a load balancing virtual machine 301, and a virtual switch that are generated after virtualization on the hardware of the computer device 300. 3022. The first virtual machine 304 and the load balancing virtual machine are 301 virtual machines created in the computer device 300.

The first virtual machine 304 is configured to send a service packet.

The virtual switch 3022 is configured to receive the service packet sent by the first virtual machine 304, and only forward the service packet to the load balancing virtual machine when the service packet needs load balancing processing. 301;

The load balancing virtual machine 301 is configured to receive the service packet forwarded by the virtual switch 3022, and forward the service packet according to a load balancing manner.

The load balancing virtual machine 301 is included in the computer device 300. When the virtual machine in the computer device 300 initiates a service and needs to perform load balancing processing, the related service message is only sent to the load balancing virtual machine 301 in the computer device 300. Perform load balancing processing. Since the load balancing related service packets are implemented in the computer device, the service packets on the plurality of computer devices are not concentrated to one load balancing node, thereby avoiding the processing delay caused by the centralized load balancing congestion. Moreover, when the computer device 300 fails, the virtual machine in the computer device 300 will also go offline, and there is no need for load balancing. Therefore, the failure of the load balancing virtual machine caused by the failure of a single computer device does not affect the load balancing virtual machine on other computer devices, and there is no active/standby switchover caused by the load balancing virtual machine failure. The business disruption.

In the embodiment of the present invention, the computer device 300 may include hardware devices such as network hardware (not shown) in addition to the central processing unit CPU 3031, the memory 3032, and the storage device 3033. The network hardware may be a switching device (such as a network card). ), routers, and/or other network devices are used to implement communication between multiple devices, and multiple devices can be connected by wireless or wired connection.

Optionally, the forwarding, by the virtual switch 3022, the service packet to the load balancing virtual machine 301 may include:

The virtual switch 3022 sends the MAC address of the load balancing virtual machine 301 to the first virtual machine 304, so that the first virtual machine 304 sends the MAC address based on the acquired MAC address of the load balancing virtual machine 301. The service packet, the virtual switch 3022 forwards the service packet to the load balancing virtual machine 301; or

The load balancing virtual machine 301 sends the MAC address to the first virtual machine 304, so that the first virtual machine 304 sends the service packet based on the obtained MAC address of the load balancing virtual machine 301. The virtual switch 3022 forwards the service packet to the load balancing virtual machine 301; or

The first virtual machine 304 sends the service packet based on the pre-configured MAC address of the load balancing virtual machine 301, and the virtual switch 3022 forwards the service packet to the load balancing virtual machine 301.

The sending, by the virtual switch 3022, the MAC address of the load balancing virtual machine 301 to the first virtual machine 304 may include:

The virtual switch 3022 receives the control packet sent by the first virtual machine 304, and returns a control response packet to the first virtual machine 304, where the control response packet includes the MAC address of the load balancing virtual machine 301. .

The control packet may be an ARP packet or an Internet Control Message Protocol (ICMP) packet, and the ICMP packet may be the sixth version of ICMPv6. Message. Taking the IPV4 protocol of the computer device 300 as an example, when the first virtual machine 304 needs to send a service packet, the first virtual machine 304 first sends an ARP packet to the virtual switch 3022, and the virtual switch 3022 performs ARP pickup to load balance the virtual The MAC address of the machine 301 is sent to the first virtual machine 304. The first virtual machine 304 combines the MAC address of the load balancing virtual machine 301 with the information such as the IP address of the load balancing virtual machine 301 that has been acquired. It is sent to the load balancing virtual machine 301 through the virtual switch 3022. In this manner, the service packet sent by the first virtual machine 304 is sent only to the load balancing virtual machine 301 in the computer device 300 when load balancing processing is required.

Referring to FIG. 4, FIG. 4 is a structural diagram of a specific implementation structure of a computer device 300 according to an embodiment of the present invention. In FIG. 4, the operating system 302 in the computer device 300 runs the virtualization layer hypervisor 3021 and the virtual switch 3022 of the virtualization management, and the virtualization layer Hypervisor 20221 performs the virtual machine (eg, the first virtual machine 304, etc.) in the computer device 300. Management, the virtual switch 3022 provides a network connection function between the virtual machines managed by the Hypervisor 3021 and between the virtual machine and the physical network, and the virtual load balancing agent unit 3023 virtualizes the load in the computer device 300 through the virtualized layer hypervisor 3021. The machine 301 is configured and managed.

The virtual load balancing agent unit 3023 is configured to receive configuration information for creating a load balancing virtual machine 301, control creation or deletion of the load balancing virtual machine 301, and control the load balancing virtual machine 301 and the virtual switch 3022. The establishment of a connection between.

The virtual load balancing agent unit 3023 may start the load balancing virtual machine 301 when the user who creates the load balancing virtual machine 301 creates the first virtual machine (for example, the first virtual machine 304) in the computer device 300. Created; when the last virtual machine created by the user who created the load balancing virtual machine 301 in the computer device 300 is deleted, the virtual load balancing agent unit 3023 instructs to delete the load balancing virtual machine 301. This can save resources of the computer device 300 and avoid waste of hardware resources caused by creating a load balancing virtual machine when the load balancing service is not required. Specifically, the virtual load balancing agent unit 3023 can refer to The implementation of the virtual load balancing proxy unit 20211 in FIG. 2 is implemented, and details are not described herein.

Optionally, the computer device 3022 further includes a second virtual machine 305, where the second virtual machine 305 is configured to process the service message of the first virtual machine 304;

When the load balancing virtual machine 301 determines that the second virtual machine 305 processes the service packet of the first virtual machine 304 by using a load balancing algorithm, the second virtual machine 305 receives the load balancing server 301. Transmitting the service packet, and sending the result of processing the service packet to the load balancing virtual machine 301;

The load balancing virtual machine 301 is further configured to send the result of processing the service packet by the second virtual machine 305 to the first virtual machine 304 through the virtual switch 3022. As an optional implementation manner, the second virtual machine 305 may directly send the result of processing the service packet to the first virtual machine 304, that is, the second virtual machine 305 does not pass the The load balancing virtual machine 301 sends the result of processing the service message to the first virtual machine 304.

The load balancing virtual machine 301 forwards the service packet of the first virtual machine 304 to the backend server (the second virtual machine 305) by using a load balancing algorithm, and sends the result of the backend server processing to the first virtual The machine 304 realizes the purpose of realizing the load balancing service on the self-device by using the load balancing virtual machine created by the computer device 300, and improves the processing efficiency of the load balancing.

Referring to FIG. 5, FIG. 5 is a schematic structural diagram of an implementation of a first computer system according to an embodiment of the present invention. The system includes a first computer device 502 and a second computer device 501, wherein the second computer device 501 includes a central processing unit CPU 50131. The first computer device 502 includes a central processing unit CPU 50231, a memory 50232, and a storage device 50233. The first computer device 502 further includes a first virtual machine generated after being virtualized on a hardware basis. 50213. The first load balancing virtual machine 50212 and the virtual switch 50222. The first virtual machine 50213 and the first load balancing virtual machine 50212 are virtual machines created in the computer device 502.

The first virtual machine 50213 is configured to send a service packet.

The first virtual switch 50222 is configured to receive the service packet sent by the first virtual machine 50213, and only forward the service packet to the first packet when the service packet needs load balancing processing. Load balancing virtual machine 50212;

The first load balancing virtual machine 50212 is configured to receive the service packet forwarded by the first virtual switch 50222, and forward the service packet according to a load balancing manner.

The first computer device 502 includes the first load balancing virtual machine 50212. When the service of the virtual machine in the first computer device 502 needs to perform load balancing processing, the related service packet is only sent to the first computer device 502. The first load balancing virtual machine 50212 performs load balancing processing. Since the load balancing related service packets are implemented in the computer device, the service packets on the plurality of computer devices are not concentrated to one load balancing node, thereby avoiding the processing delay caused by the centralized load balancing congestion. Moreover, when the first computer device 502 fails, the virtual machine in the first computer device 502 is also offline, and there is no need for load balancing. Therefore, the load balancing virtual machine failure caused by the failure of a single computer device does not affect the load balancing virtual machine on other computer devices, and there is no service interruption caused by the load balancing virtual machine failure. .

In the embodiment of the present invention, the second computer device 501 may include hardware devices such as network hardware (not shown) in addition to the central processing unit CPU 50131, the memory 50132, and the storage device 50133; the network hardware may be exchanged. Devices (network cards, etc.), routers, and/or other network devices are used to implement communication between multiple devices, and multiple devices can be connected by wireless or wired connection.

Optionally, the first virtual switch 50222 to forward the service packet to the first load balancing virtual machine 50212 may include:

The first virtual switch 50222 sends the MAC address of the first load balancing virtual machine 50212 to the first virtual machine 50213, so that the first virtual machine 50213 is based on the acquired first load balancing virtual machine. The 50212 MAC address sends the service packet, and the first virtual switch 50222 forwards the service packet to the first load balancing virtual machine 50212; or

The first load balancing virtual machine 50212 sends its MAC address to the first virtual machine 50213, so that the first virtual machine 50213 sends the service packet based on the first load balancing virtual machine 50212 MAC address. The first virtual switch 50222 forwards the service packet to the first load balancing virtual machine 50212; or

The first virtual machine 50213 sends the service packet based on the pre-configured MAC address of the first load balancing virtual machine 50212, and the first virtual switch 50222 forwards the service packet to the first Load balancing virtual machine 50212.

The sending, by the first virtual switch 50222, the MAC address of the first load balancing virtual machine 50212 to the first virtual machine 50213 includes:

The first virtual switch 50222 receives the control packet sent by the first virtual machine 50213, and returns a control response message to the first virtual machine 50213, where the control response message includes the first load balancing virtual machine. 50212's MAC address.

The control packet can be an ARP packet or an ICMPv6 packet. The first virtual machine 50214 first sends an ARP packet to the first virtual switch 50222, and the first virtual switch 50222 performs the ARP generation. The MAC address of the load balancing virtual machine 301 is sent to the first virtual machine 50213. The first virtual machine 50213 combines the acquired first load balancing virtual machine 50212 according to the MAC address of the first load balancing virtual machine 50212. The information such as the IP address is sent to the first load balancing virtual machine 50212 through the first virtual switch 50222. In this manner, the service packet sent by the first virtual machine 50213 is sent only to the first load balancing virtual machine 50212 in the first computer device 502 when load balancing processing is required.

Referring to FIG. 6, FIG. 6 is a structural diagram of a specific implementation structure of a computer system in a first embodiment of the present invention. In FIG. 6, the operating system 5022 in the second computer device 502 runs the virtualization layer hypervisor 50221 and the first virtual switch 50222 of the virtualization management, and the virtualization layer Hypervisor 50221 pairs the virtual machine in the first computer device 502 (for example, the first The virtual machine 50213 or the like performs management of the first virtual switch 50222 managed by the Hypervisor 50221. The network connection function is provided between the virtual machines and between the virtual machine and the physical network, and the virtual load balancing agent unit 50111 configures and manages the first load balancing virtual machine 50212 in the first computer device 502 through the virtual machine layer Hypervisor 50221. The second computer device 501 functions as a cloud management platform in the computer system, and its operating system 5012 runs a virtualization layer hypervisor 50121 of virtualization management. The second computer device 501 further includes a virtual machine management unit 50112 and a virtual load balancing management unit 50111. The virtual machine management unit 50112 implements configuration and management of a virtual machine in the second computer device 502. The virtual load balancing management unit 50111 The configuration information for creating a load balancing virtual machine or the configuration information of the virtual machine is created, and the related configuration information is saved, and the configuration information for creating the load balancing virtual machine is saved in the virtual load balancing data device 50114. The first computer device 502 also includes a first virtual load balancing agent unit 50211.

The virtual load balancing management unit 50111 is configured to receive configuration information for creating the first load balancing virtual machine 50212, and send the configuration information of the first load balancing virtual machine 50212 to the first virtual load balancing. Proxy unit 50211;

The first virtual load balancing agent unit 50211 is configured to receive the configuration information of the first load balancing virtual machine 50212, and instruct the virtual machine management unit to create the configuration information according to the first load balancing virtual machine 50212. The first load balancing virtual machine 50212.

The first virtual load balancing agent unit 50211 may indicate when the user who created the first load balancing virtual machine 50212 creates the first virtual machine (for example, the first virtual machine 50213) in the first computer device 502. The virtual machine management unit 50112 creates the first load balancing virtual machine 50212; when the last virtual machine created by the user who created the first load balancing virtual machine 50212 in the first computer device 502 is deleted, The first virtual load balancing proxy unit 50211 instructs the virtual machine management unit 50112 to delete the first load balancing virtual machine 50212. This can save resources of the first computer device 502 and avoid waste of hardware resources caused by creating a load balancing virtual machine when the load balancing service is not required. Specifically, the first virtual load balancing proxy unit 50211 can refer to the virtual load in FIG. 2 above. The implementation of the balancing agent unit 20211 is implemented and will not be described again.

In this embodiment, the first virtual load balancing proxy unit 50211 is further configured to instruct the virtual machine management unit 50112 to configure a first virtual network card for the first load balancing virtual machine 50212 and establish the first virtual network card. A connection with the first virtual switch 50222.

The virtual machine management unit 50112 establishes that the connection between the virtual network card and the first virtual switch 50222 can be implemented by configuring a network identifier of the first virtual switch 50222. For example, the first virtual switch 50222 has three network interfaces, and the network identifier is 100. When the first load balancing virtual machine 50212 and the virtual network card of the first virtual machine 50213 have an interface with the network identifier of 100, the first virtual The machine 50213 can communicate with the interface of the load balancing virtual machine 20212 with the network identifier of 100 through an interface with a network identifier of 100. The virtual machine management unit 50112 may configure a network identifier for the first virtual switch 50222 according to the first virtual load balancing proxy unit 50211 or the virtual load balancing management unit 50111.

In the embodiment of the present invention, the first virtual machine 50213 is a virtual machine created by a user who creates configuration information of the first load balancing virtual machine 50212; the virtual load balancing management unit 50111 receives the first virtual machine 50213. Configuration information, and when it is determined that the first virtual machine 50213 is the first virtual machine created by the user on the first computer device 502, notifying the information that the first virtual machine 50213 is created a first virtual load balancing agent unit 50211; the first virtual load balancing agent unit 50211 instructing the virtual machine management unit when receiving the notification of creating the first virtual machine 50213 sent by the virtual load balancing management unit The first load balancing virtual machine 50212 is created 50112. In this way, the resources of the computer device 502 can be saved, and the occupation and waste of hardware resources caused by the creation of the load balancing virtual machine are avoided when the load balancing service is not required. Similarly, when the user who creates the configuration information of the first load balancing virtual machine 50212 is deleted when the last virtual machine in the computer device 502 is deleted, the virtual load balancing management unit 50111 notifies the first virtual load balancing. The proxy unit 50211, the first virtual load balancing proxy unit 50211 instructs the virtual machine management unit 50112 to create the first load balancing virtual machine 50212. This also saves resources of the computer device 502 and avoids When load balancing services are not required, hardware resources are occupied and wasted due to load balancing virtual machines.

The above is the virtual load balancing management unit 50111 notifying the first virtual load balancing agent unit 50211 to instruct the virtual machine management unit 50112 to create or delete the first load balancing virtual machine 50212. The virtual load balancing management unit 50111 notifies the virtual machine management unit 50112 to create or delete the first load balancing virtual machine 50212, and details are not described herein.

Optionally, the first computer device 502 further includes a second virtual machine 50214, where the second virtual machine 50214 is configured to process the service message of the first virtual machine 50213; When the equalization virtual machine 50212 determines that the second virtual machine 50214 processes the service packet of the first virtual machine 50213 by using a load balancing manner, the second virtual machine 50214 receives the first load balancing virtual machine 50212 for forwarding. The service packet, and the result of processing the service packet is sent to the first load balancing virtual machine 50212; the first load balancing virtual machine 5021 processes the service by the second virtual machine 50214 The result of the message is sent to the first virtual machine 50213 through the first virtual switch 50222.

The first load balancing virtual machine 50212 forwards the service packet of the first virtual machine 50213 to the backend server (the second virtual machine 50214) through a load balancing algorithm, and sends the result of the backend server processing to the first server. The first virtual machine 50213 implements the purpose of realizing the load balancing service on the device by using the load balancing virtual machine created by the computer device 502, and improves the processing efficiency of the load balancing.

The first computer system of the embodiments of the present invention may further include a third computer device. As shown in FIG. 7, the third computer device 503 is similar to the computer device 502, and includes a second virtual load balancing proxy unit 50311, a second virtual switch 50322, and a third virtual machine 50313;

The second virtual load balancing proxy unit 50312 is configured to receive the configuration information of the second load balancing virtual machine 50312 sent by the virtual load balancing management unit 50111, and create configuration information of the second load balancing virtual machine 50312 based on the configuration information. Instructing the virtual machine management unit 50112 to create the second load balancing virtual machine 50312, and instructing the virtual machine management unit The second virtual network card is configured for the second load balancing virtual machine 50312 and establishes a connection between the second virtual network card and the second virtual switch 50322.

The second virtual switch 50322 is configured to receive the second service packet initiated by the third virtual machine 50313. When the second service packet needs to be load balanced, the second service packet is forwarded only to the second service packet. The second load balancing virtual machine 50312;

The second load balancing virtual machine 50312 is configured to receive the second service packet forwarded by the second virtual switch 50322, and forward the second service packet according to a load balancing manner.

Correspondingly, the virtual machine management unit 50112 may configure a first IP address and a first MAC address for the first load balancing virtual machine 50212, and configure a second IP address and a second for the second load balancing virtual machine. MAC address. The first IP address may be the same as the second IP address, and the first MAC address may also be the same as the second MAC address, so that the IP address resource in the network formed by the computer device may be saved or MAC address resource to improve the utilization of IP address resources or MAC address resources.

The embodiments shown in FIG. 5, FIG. 6, and FIG. 7 can also be implemented by referring to the specific implementation manner in FIG. 2, and details are not described herein again.

Optionally, the embodiment of the present invention further provides a second computer system, including a first computer device (such as computer device 202 in FIG. 2 or computer device 502 in FIG. 7), and a second computer device (as shown in FIG. 2). Computer device 201 or computer device 501 in FIG. 7 and a third computer device (such as computer device 203 in FIG. 2 or computer device 503 in FIG. 7), wherein the first computer device, the second computer device And the third computer device respectively include a central processing unit CPU, a memory and a storage device, the second computer device configuring and managing the virtual machine on the first computer device and the third computer device, the first The computer device includes a first virtual switch, and the third computer device includes a second virtual switch;

The second computer device is configured to receive configuration information for creating a load balancing virtual machine on the first computer device, create a first load balancing virtual machine in the first computer device, and establish the first load Balancing the connection of the virtual machine to the first virtual switch; and receiving Creating configuration information of the load balancing virtual machine on the third computer device, creating a second load balancing virtual machine in the third computer device, and establishing the second load balancing virtual machine and the second virtual switch Connection;

The second computer system includes the first computer device, the second computer device, and the third computer device, and the second computer device respectively creates a first load balancing virtual machine in the first computer device, where The second load balancing virtual machine is created in the third computer device, so that each load balancing virtual machine in each computer device processes the load balancing service in the respective computer device, and the service message sent by the virtual machine in each computer device is only The load balancing virtual machine in the computer device is used for load balancing processing. In this way, service packets on multiple computer devices are not concentrated into one load balancing node, which avoids the processing delay caused by centralized load balancing congestion. Moreover, when a computer device fails, the virtual machine in the computer device will also go offline, and there will be no need for load balancing. Therefore, the load balancing virtual machine failure caused by the failure of a single computer device does not affect the load balancing virtual machine on other computer devices, and there is no service interruption caused by the load balancing virtual machine failure. .

The second computer system, the first virtual switch forwarding the first service packet to the first load balancing virtual machine includes:

Sending, by the first virtual switch, the MAC address of the first load balancing virtual machine to the first virtual machine, so that the first virtual machine sends the first load balancing virtual machine MAC address based on the acquired The first virtual switch forwards the service packet to the service packet The first load balancing virtual machine; or

The first load balancing virtual machine sends its MAC address to the first virtual machine, so that the first virtual machine sends the service packet based on the first load balancing virtual machine MAC address, the first The virtual switch forwards the service packet to the first load balancing virtual machine; or

Correspondingly, the sending, by the first virtual switch, the MAC address of the first load balancing virtual machine to the first virtual machine includes:

Receiving, by the first virtual switch, a control packet sent by the first virtual machine, and returning a control response packet to the first virtual machine, where the control response packet includes a MAC address of the first load balancing virtual machine. . The control packet can be an ARP packet or an ICMPv6 packet.

Optionally, the first computer device further includes a second virtual machine, where the second virtual machine is configured to process the service packet of the first virtual machine;

In the embodiment of the present invention, the IP address of the first load balancing virtual machine is the same as the IP address of the second load balancing virtual machine; or the MAC address of the first load balancing virtual machine and the second load balancing The virtual machine has the same MAC address. In this way, the address resource of the IP address or the MAC address in the second computer system can be saved, and the resource utilization rate can be improved.

Referring to FIG. 8, FIG. 8 is a schematic flowchart of a method for implementing load balancing according to an embodiment of the present invention. As shown in FIG. 8, the method includes:

Step 800: The second computer device receives configuration information for creating the load balancing virtual machine on the first computer device, where the first computer device includes a central processing unit CPU, a memory, and a storage device, and the first computer device further includes The first virtual machine and the virtual switch;

Step 802: The second computer device creates a load balancing virtual machine on the first computer device according to the configuration information, and establishes a connection between the load balancing virtual machine and a virtual switch in the first computer device. ;

Step 804: The virtual switch receives the service packet sent by the first virtual machine, and only forwards the service packet to the load balancing virtual machine when the service packet needs to be load balanced.

Step 806: The load balancing virtual machine receives the service packet forwarded by the first virtual switch, and forwards the service packet according to a load balancing manner.

In the above method, the first computer device includes a load balancing virtual machine, and when the virtual machine in the first computer device initiates a service, and the load balancing process is required, the related service packet is only sent to the load in the first computer device. Balance the virtual machine for load balancing. Since the load balancing related service packets are implemented in the computer device, the service packets on the plurality of computer devices are not concentrated to one load balancing node, thereby avoiding the processing delay caused by the centralized load balancing congestion. Moreover, when the first computer device fails, the virtual machine in the first computer device will also go offline, and there is no need for load balancing. Therefore, the load balancing virtual machine failure caused by the failure of a single computer device does not affect the load balancing virtual machine on other computer devices, and there is no service interruption caused by the load balancing virtual machine failure. .

Optionally, in the foregoing step 804, the virtual switch only forwards the service packet to the load balancing virtual machine, including:

The virtual switch sends the MAC address of the load balancing virtual machine to the first virtual machine, so that the first virtual machine sends the service packet based on the obtained load balancing virtual machine MAC address. The virtual switch forwards the service packet to the load balancing virtual Machine; or,

The sending, by the virtual switch, the MAC address of the load balancing virtual machine to the first virtual machine may include:

The virtual switch receives the control packet sent by the first virtual machine, and returns a control response packet to the first virtual machine, where the control response packet includes a MAC address of the load balancing virtual machine. The control packet can be an ARP packet or an ICMPv6 packet.

Before the second computer device creates a load balancing virtual machine on the first computer device according to the configuration information, the method further includes:

The first computer device receives configuration information for creating a load balancing virtual machine, instructing the second computer device to create the load balancing virtual machine.

In the embodiment of the present invention, the second computer device further includes a second virtual machine, where the second virtual machine has the capability of processing the service packet of the first virtual machine, and the method further includes:

The second virtual machine and the first virtual machine are in one computer device, which can save industry It can handle the bandwidth occupation caused by cross-computer equipment and improve the efficiency of business processing.

The implementation of the method embodiment in the embodiment of the present invention may also be implemented by referring to the implementation manner of the embodiment shown in FIG. 2, and details are not described herein.

In the following, in conjunction with the application scenario shown in FIG. 2, the user applies for the virtual machine 20213 in the computer device 202 through the cloud management platform, and implements a specific service (for example, accessing a video service) through the virtual machine 20213, and the service may be performed by the computer device 202. The virtual machine 20214 or the virtual machine 20314 in the computer device 203 is processed by the back-end server, and the load balancing of the service initiated by the virtual machine 20213 between the virtual machine 20214 and the virtual machine 20314 is required as an example, which is implemented in the embodiment of the present invention. The implementation method of the load balancing service is described in detail.

Figure 9 is a schematic flow chart of the implementation process. It should be noted that the implementation process shown in FIG. 9 is still described by using a virtual machine with load balancing function as a load balancer. For implementing the load balancer in a container or a namespace, the implementation of the virtual machine as a load balancer can be implemented as follows.

As shown in FIG. 9, a method for implementing load balancing services according to an embodiment of the present invention includes:

Step 900: The computer device 201 acquires configuration information for creating a load balancing virtual machine in the computer device 202.

Step 902: The computer device 201 creates a load balancing virtual machine 20212 in the computer device 202 according to the acquired configuration information of the load balancing virtual machine.

Step 904: Configure a forwarding mode of the load balancing service in the computer device 202, so that the service initiated by the virtual machine 20213 is only forwarded to the load balancing virtual machine 20212 created in the computer device 202 for load balancing processing;

The service initiated by the virtual machine 20213 is a load balancing service, and the load balancing service is a service that needs to perform load balancing processing.

Step 906: The load balancing virtual machine 20212 receives the service sent by the virtual machine 20213, and selects a backend server to process the service initiated by the virtual machine 20213.

Step 908: The backend server returns the processed service initiated by the virtual machine 20213. Give virtual machine 20213.

In the above step 900, the computer device 201, as the cloud management platform, can obtain the configuration information of the load balancing virtual machine that needs to be created, for example, the load balancing virtual machine delivered by the user through the software related to the cloud management platform running on the computer device 201. Configuration information, or configuration information about the load balancing virtual machine pre-stored in the computer device 201. Moreover, the computer device 201, as a cloud management platform, may also acquire configuration information of other service virtual machines, such as the virtual machine 20213 in the computer device 202. The manner in which the computer device 201 obtains the configuration information of the service virtual machine includes receiving configuration information delivered by the user through the cloud management platform related software or configuration information pre-stored in the computer device 201.

The embodiment of the present invention is that the user requests the virtual machine 20213 in the computer device 202 through the cloud management platform, and the service request initiated by the virtual machine 20213 needs to perform load balancing processing. In order to avoid the congestion and the like caused by the centralized load balancing node in the prior art, the embodiment of the present invention creates a load balancing virtual machine 20212 in the computer device 202, and the load balancing virtual machine 20212 processes the need initiated by the virtual machine 20213. Service request for load balancing processing. Therefore, in step 900, the computer device 201 needs to acquire the configuration information of the load balancing virtual machine created in the computer device 202, that is, the configuration information of the load balancing virtual machine 20212, and according to the acquired configuration information of the load balancing virtual machine. A load balancing virtual machine is configured in the computer device 202.

In the embodiment of the present invention, the configuration information of the load balancing virtual machine includes, but is not limited to, an identifier of the load balancing virtual machine 20212, and a VIP address of the load balancing virtual machine 20212. The VIP of the load balancing virtual machine 20212 indicates a load balancing service. For example, when the IP address of the load balancing virtual machine 20212 is configured as a VIP, the load balancing virtual machine 20212 can be addressed as a node that handles load balancing services. Optionally, in the embodiment of the present invention, the configuration information of the load balancing virtual machine further includes information about a backend server related to the load balancing virtual machine 20212.

Optionally, the configuration information may further include virtual machine parameter information, where the virtual machine parameter information includes, but is not limited to, the number of virtual central processing units, the size of the virtual memory, and the load. The information such as the logical network where the virtual machine 20212 is located is equalized. The parameter information of the virtual machine may be a pre-configured parameter list, and different parameter lists may be selected according to different requirements. The parameter list may be specified when the user sends the configuration information, or may be obtained by the computer device 201 in the stored configuration information list according to the previous policy. The embodiment of the present invention does not limit the virtual machine parameter of the load balancing virtual machine 20212. The way information is.

In step 902, the computer device 201 creates a load balancing virtual machine 20212 according to the acquired configuration information of the load balancing virtual machine.

As a cloud management platform, the computer device 201 can create a normal virtual machine in the computer device 202, and can also create a load balancing virtual machine. The computer device 201 can create a load balancing virtual machine 20212 in the computer device 202 based on the acquired configuration information of the load balancing virtual machine 20212.

Optionally, the foregoing method may further include the step 901: the computer device 201 creates the virtual machine 20213 in the computer device 202. The virtual machine 20213 is a service virtual machine created by a user, and the virtual machine 20213 is a virtual machine that initiates a service. The service initiated by the virtual machine 20213 includes a load balancing service, that is, a load balancing service initiated by the virtual machine 20213, which is a service that needs to perform load balancing processing.

When the user who created the load balancing virtual machine configuration information creates the first virtual machine on the computer device 202, for example, when the first virtual machine 20213 that initiates the load balancing service is created, the computer device 201 creates the load balancing virtual machine 20212. When the last virtual machine created by the user is deleted, for example, the last virtual machine 20213 that initiated the load balancing service is deleted, the computer device 201 deletes the load balancing virtual machine 20212. In this way, the load balancing virtual machine 20212 is created only when the user creates the virtual machine that initiates the load balancing service, which can avoid the problem that the resources of the computer device 202 caused by the creation of the load balancing virtual machine are occupied when the user does not create the virtual machine. The resource utilization of computer device 202 is increased.

The computer device 201 determines whether the created virtual machine is the first service virtual machine created by the user who has configured the virtual load balancing service, and may use the information according to the user, for example, The identification information of the user, etc., determines whether it is the first service virtual machine created by the user who has configured the virtual load balancing service.

Similarly, when it is determined that the last service virtual machine created by the user who has configured the virtual load balancing service is deleted, the user information may be determined according to the information of the user, and details are not described herein.

In step 902, the computer device 201 creates a load balancing virtual machine 20212 according to the acquired configuration information of the load balancing virtual machine, and further includes the computer device 201 as the load balancing virtual network according to the network information in the configuration information of the load balancing virtual machine. The machine 20212 configures a virtual network card.

In the above step 904, a forwarding mode in which the load balancing service is configured in the computer device 202 by the computer device 201 may be used. The forwarding mode forwards the service packet of the load balancing service initiated by the virtual machine that initiates the service on the computer device 202, such as the virtual machine 20213, to the load balancer.

The computer device 201 configures the forwarding mode of the load balancing service in the computer device 202, which can be implemented by configuring the forwarding mode of the virtual switch in the computer device 202. The virtual switch in the computer device 202 is a virtual switch 20222. The forwarding mode of the virtual switch 20222 is configured to perform a directional forwarding process on the service packet that needs to be load-balanced and sent by the virtual machine 20213. The service message sent for load balancing processing is sent only to the load balancing virtual machine 20212 in the computer device 202. Optionally, the forwarding mode of the virtual switch 20222 may be an ARP or NDP pickup mode, that is, the virtual switch 20222 performs directional forwarding of the service packet by using an ARP or NDP pickup mode.

When the virtual machine 20213 initiates a service packet of the load balancing service, the virtual machine 20213 sends the service packet of the load balancing service to the load balancing virtual machine 20212 according to the forwarding mode of the load balancing service. Specifically, when the virtual machine 20213 initiates a load balancing service that needs to perform load balancing processing, the control device may first send a control packet with the VIP as the destination IP, and the control packet may be a broadcast packet of the load balancing service. In the normal case, the virtual machine 20213 initiates a control packet with the VIP as the destination IP, and is forwarded to all IP addresses. A load balancing node (such as a load balancer, etc.) that is a VIP. However, in the embodiment of the present invention, since the computer device 201 configures the forwarding mode of the load balancing service in the computer device 202, the virtual switch 20222 in the computer device 202 only balances the load after receiving the control packet sent by the virtual machine 20213. The MAC address of the virtual machine 20212 is sent to the virtual machine 20213. The virtual machine 20213 sends the service packet of the load balancing service to the load balancing virtual machine 20212 according to the obtained MAC address of the load balancing virtual machine 20212. The services initiated by the virtual machine 20213 are only forwarded to the load balancing virtual machine 20212 created in the computer device 202 for load balancing processing.

Optionally, before the virtual machine 20213 sends the service packet, the control packet sent to the virtual switch 20222 may be an ARP packet or an ICMP packet.

The embodiment of the present invention uses a cloud platform supporting OpenStack as an example to describe the implementation manner of the ARP pickup technology. When the virtual machine 20213 first initiates the service of the load balancing process, the control packet is sent to the virtual switch 20222, for example, an ARP request packet is sent to the virtual switch 20222, and the virtual switch 20222 performs ARP pickup. When the virtual machine 20213 initiates the load balancing processing service for the first time, the virtual machine 20213 does not know the MAC address of the load balancing virtual machine that processes the load balancing service, and only knows the VIP address of the load balancing service. Therefore, when the virtual machine 20213 initiates the load balancing processing service for the first time, it first sends an ARP request to obtain the MAC address of the load balancing virtual machine that processes the load balancing service.

The destination IP address of the ARP request packet is a VIP, and the VIP is an IP address that the load balancing virtual machine can access. The destination MAC address of the ARP request packet may be a MAC address of a general load balancing virtual machine. The MAC address is not a MAC address of a specific load balancing virtual machine, and the destination MAC address may also be empty. The ARP pickup technology is implemented by a flow table. The flow table technology generally consists of two parts, one is the matching condition and the other is the action. The matching condition of the flow table of the ARP pickup is: the protocol type is ARP, the target protocol address of the ARP For the VIP, the VIP is the IP address that the load balancing virtual machine can be accessed, and the ARP operation type is the ARP request. When the virtual switch 20222 performs ARP pickup on the ARP request sent by the virtual machine 20213, the load balancing virtual machine 20212 The MAC address is returned to the virtual machine 20213.

After the virtual switch 20222 sends the ARP reply packet to the virtual machine 20213, the virtual machine 20213 sends a service packet that needs load balancing processing. The IP address of the service packet is a VIP address, and the MAC address is an ARP response from the virtual switch 20222. The MAC of the load balancing virtual machine 20212 obtained in . The virtual machine 20213 sends the service packet to the virtual switch 20222. After receiving the service packet sent by the virtual machine 20213, the virtual switch 20202 sends the service packet to the load balancing virtual machine 20212 according to the MAC address carried in the service packet.

The virtual switch 20222 receives the service packet sent by the virtual machine 20213, and sends the service packet to the load balancing virtual machine 20212 according to the IP address and the MAC address of the service packet. The load balancing virtual machine (such as the load balancing virtual machine 20312) on the other computer device (such as the computer device 203) does not receive the service packet of the virtual machine 20213, and does not process the service packet that needs to be load balanced and sent by the virtual machine 20213. . The purpose of realizing the service message of the virtual machine in each computer device is only implemented on the load balancing virtual machine in the computer device when load balancing processing is required.

As an optional implementation manner, the forwarding mode of the load balancing service configured by the computer device 201 on the computer device 202 can also be implemented by:

The computer device 201 instructs the load balancing virtual machine 20212 to send the MAC address of the load balancing virtual machine 20212 to the virtual machine (eg, the virtual machine 20213) that initiates the service; or

The computer device 201 configures the MAC address of the load balancing virtual machine 20212 on the virtual machine (eg, virtual machine 20213) that initiates the service.

When the computer device 201 instructs the load balancing virtual machine 20212 to send the MAC address of the load balancing virtual machine 20212 to the virtual machine (eg, the virtual machine 20213) that initiates the service, the virtual machine that initiates the service ( For example, the virtual machine 20213) receives the MAC address of the load balancing virtual machine 20212 as the destination MAC address when initiating the load balancing service. When the computer device 201 is in the virtual machine that initiates the service (for example, the virtual machine 20213) When the MAC address of the load balancing virtual machine 20212 is configured, the MAC address of the load balancing virtual machine 20212 is the destination MAC address when the virtual machine (for example, the virtual machine 20213) that initiates the service sends the load balancing service.

In the foregoing step 906, the load balancing virtual machine 20212 receives the service sent by the virtual machine 20213, and selects a backend server to process the service initiated by the virtual machine 20213. The method may include the following steps: the load balancing virtual machine 20212 receives the service. After the packet is received, it is first determined whether a service is sent to the service server by the backend server. If the backend server has been allocated, the service packet is sent to the already assigned backend server. If a backend server has not been assigned, select a backend server to serve it through a load balancing algorithm. The load balancing algorithms include, but are not limited to, polling, weighted polling, random, minimum or source-to-ground hashing, location-based weighted polling, and the like. The location-prioritized weighted polling mode may preferentially select the back-end server in the computer device where the virtual machine that originated the service message is located, that is, the virtual machine 20214 in the computer device 202 as the back-end server that processes the virtual machine 20213 service. This can save the speed of the back-end server processing service packets and save network bandwidth. Of course, in step 906, the load balancing virtual machine 20212 may also process the service request of the virtual machine 20213 through the load balancing algorithm, and then send it to the virtual machine 20314 in the computer device 203, and the virtual machine 20314 serves as the processing virtual machine 20213. The backend server processes the services initiated by the virtual machine 20213.

It should be noted that the dotted line of step 906 in FIG. 9 indicates that the load balancing virtual machine 20212 can forward the service initiated by the virtual machine 20213 to the backend server in the computer device 202, that is, the virtual machine 20214, and can also forward the service. The backend server and virtual machine 20314 in the computer device 203 are provided.

In the foregoing step 908, the manner in which the backend server returns the processed service request to the virtual machine 20213 may include two types:

Method 1: The backend server processing the virtual machine 20213 service request, and returning the processed result to the virtual machine 20213 through the load balancing virtual machine 20212;

Manner 2: The back-end server that processes the service request of the virtual machine 20213 does not pass the load balancing virtual machine 20212, and directly returns the processed result to the virtual machine 20213.

In the foregoing manner 1, the load balancing virtual machine 20212 writes the source address (including the IP address and the MC address, etc.) of the service request as load balancing when forwarding the service request of the virtual machine 20213 to the backend server that processes the virtual machine 20213 service request. The address of the virtual machine 20212 writes the destination address as the address of the backend server that processes the virtual machine 20213 service request. In this way, the backend server processing the service request of the virtual machine 20213 returns the result of processing the service request to the load balancing virtual machine 20212 according to the source address carried in the service request, and the load balancing virtual machine 20212 returns the service request received. Sent to virtual machine 20213.

In the foregoing manner 2, the load balancing virtual machine 20212 writes the source address (including the IP address, the MC address, and the like) as the address of the virtual machine 20213 when forwarding the service request of the virtual machine 20213 to the backend server that processes the virtual machine 20213 service request. The destination address is written as the address of the backend server that processes the virtual machine 20213 service request. In this way, the backend server that processes the virtual machine 20213 service request returns the result of processing the service request to the virtual machine 20213 according to the source address carried when receiving the service request.

It should be noted that the dotted line of step 908 in FIG. 9 indicates that the load balancing virtual machine 20212 forwards the service initiated by the virtual machine 20213 to the backend server in the computer device 202, that is, the virtual machine 20214. The processed service request is returned by the virtual machine 20214 to the virtual machine 20213. When the load balancing virtual machine 20212 forwards the service initiated by the virtual machine 20213 to the backend server in the computer device 203, that is, the virtual machine 20314, the virtual machine 20314 returns the processed service request to the virtual machine 20213.

In the embodiment of the present invention, the computer device 202 and the computer device 203 are also referred to as service computer devices. A virtual machine running on computer device 202 and/or computer device 203, such as virtual machine 20213, is also referred to as a virtual machine that initiates a service, and the services initiated by the computer device 202 include load balancing services.

The implementation process of the above steps 900- 908 is described by taking the computer device 201 as the cloud management platform to create the load balancing virtual machine 20212 in the computer device 202 as an example. It can be understood that the process for the computer device 201 to create the load balancing virtual machine 20312 in the computer device 203 is similar to the above process and will not be described again.

Thus, computer device 201 can create a load balancing virtual machine in computer device 202 and computer device 203, respectively. The service packets of the load balancing service initiated by the service virtual machine (for example, the virtual machine 20213 or the virtual machine 20313) on each of the business computer devices (for example, the computer device 202 and the computer device 203) are forwarded to the local load balancing virtual machine. That is, the load balancing service initiated by the service virtual machine 20213 in the business computer device 202 is only forwarded to the load balancing virtual machine (ie, the load balancing virtual machine 20212) local to the computer device 202 to implement the load balancing service. The load balancing service initiated by the service virtual machine 20313 in the service computer device 203 is only forwarded to the load balancing virtual machine (ie, the load balancing virtual machine 20312) local to the computer device 203 to implement the load balancing service.

The load balancing service initiated by the service virtual machine 20213 in the service computer device 202 is only forwarded to the load balancing virtual machine (ie, the load balancing virtual machine 20212) local to the computer device 202 to implement the load balancing service; in the service computer device 203 The load balancing service initiated by the service virtual machine 20313 is only forwarded to the load balancing virtual machine (ie, the load balancing virtual machine 20312) local to the computer device 203 to implement the load balancing service. Therefore, the IP address of the load balancing virtual machine 20212 in the business computer device 202 can be the same as the IP address of the load balancing virtual machine 20312 in the business computer device 203; likewise, the MAC address of the load balancing virtual machine 20212 in the business computer device 202. The MAC address of the load balancing virtual machine 20312 in the business computer device 203 can be the same. In this way, IP address resources and/or MAC address resources in the network formed by the computer devices can be saved.

Those of ordinary skill in the art will appreciate that the elements and algorithm steps of the various examples described in connection with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both, for clarity of hardware and software. Interchangeability, the composition and steps of the various examples have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the solution. Professional skill worker Different methods may be used to implement the described functionality for each particular application, but such implementation should not be considered to be beyond the scope of the present invention.

A person skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working process of the system, the device and the unit described above can refer to the corresponding process in the foregoing method embodiment, and details are not described herein again.

In the several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods may be implemented in other manners. For example, the device embodiments described above are merely illustrative. For example, the division of the unit is only a logical function division. In actual implementation, there may be another division manner, for example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, or an electrical, mechanical or other form of connection.

The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the embodiments of the present invention.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a standalone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention contributes in essence or to the prior art, or all or part of the technical solution may be embodied in the form of a software product stored in a storage medium. Including a number of instructions for causing a computer device (which may be a personal computer, server, or network device, etc.) to perform the methods of the various embodiments of the present invention All or part of the steps. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and the like. .

The above is only the specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any equivalent person can be easily conceived within the technical scope of the present invention by any person skilled in the art. Modifications or substitutions are intended to be included within the scope of the invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

A method for implementing load balancing is applied to a computer system including at least two computer devices, wherein each computer device includes a central processing unit and a memory, and the at least two computer devices include a first computer device and a second computer The device, the first computer device is a cloud management platform, the second computer device is running a virtual machine that initiates a service, and the service is a load balancing service, wherein

The first computer device obtains configuration information of a load balancer to be created in the second computer device, the configuration information of the load balancer includes an identifier of the load balancer, and a virtualized IP of the load balancer Address VIP, the VIP of the load balancer indicates the load balancing service;

The first computer device creates the load balancer on the second computer device according to the configuration information;

The first computer device configures a forwarding mode of the load balancing service on the second computer device, the forwarding mode causing the load balancing initiated by the virtual machine initiating the service on the second computer device The service packet of the service is forwarded to the load balancer.

The virtual machine that initiates the service on the second computer device sends the service packet to the load balancer according to the forwarding mode;

After receiving the service packet, the load balancer on the second computer device selects at least one backend server to execute the service, and the backend server is configured to execute the load balancing service.
The method of claim 1, wherein the configuring, by the first computer device, the forwarding mode of the load balancing service on the second computer device comprises:

The first computer device configures a virtual switch deployed on the second computer device, and sends a MAC address of the load balancer to the virtual machine that initiates the service;

The sending the service packet to the load balancer according to the forwarding mode of the virtual machine that initiates the service on the second computer device includes:

The virtual machine that sends the service sends the MAC address of the load balancer as the service packet of the destination address, and the virtual switch sends the service packet to the load according to the destination address of the service packet. Equalizer.
The method according to claim 2, wherein before the virtual machine that initiates the service sends the MAC address of the load balancer as the service message of the destination address, the method further includes:

The virtual machine that initiates the service sends a control packet to the virtual switch, where the control packet carries the VIP;

The virtual switch sends a response message of the control packet to the virtual machine that initiates the service, where the response message carries the MAC address of the load balancer.
The method according to claim 3, wherein the control message is an address resolution protocol ARP message or an internet control message protocol ICMP message.
The method according to claim 1, wherein the first computer device configures a forwarding mode of the load balancing service on the second computer device, including:

The first computer device instructs the load balancer to send a MAC address of the load balancer to the virtual machine that initiates the service; or

The first computer device configures a MAC address of the load balancer on the virtual machine that initiates the service;

Correspondingly, the sending, by the virtual machine of the initiating service on the second computer device, the sending the service packet to the load balancer according to the forwarding mode comprises:

The virtual machine that initiates the service sends the MAC address of the load balancer as the service packet of the destination address, and the virtual switch deployed on the second computer device sends the service according to the destination address of the service packet. The message is sent to the load balancer.
The method according to any one of claims 1-5, wherein the creating, by the first computer device, the load balancer on the second computer device according to the configuration information comprises:

After the first computer device creates the first virtual machine that initiates the load balancing service on the second computer device, the load equalizer is created.
The method according to any one of claims 1 to 6, wherein the method further comprises:

And deleting, by the first computer device, the created load balancer after the last one created by the second computer device initiates the load balancing service is deleted.
The method according to any one of claims 1 to 7, wherein the method further comprises:

The selected backend server directly returns the result of executing the service to the virtual machine that initiates the service; or

The selected backend server returns an execution result of the service to the load balancer, and the load balancer returns the execution result to the virtual machine that initiates the service.
A method for implementing load balancing is applied to a computer system including at least three computer devices, wherein each computer device includes a central processing unit and a memory, and the at least three computer devices include a first computer device, the a computer device as a cloud management platform, wherein at least three computer devices other than the first computer device are used as business computer devices, and each business computer device runs a virtual machine that initiates a service, and the service For load balancing services, it is characterized by

The first computer device respectively obtains configuration information of a load balancer to be created on each service computer device, where configuration information of the load balancer includes an identifier of the load balancer, and virtualization of the load balancer IP address VIP, the VIP of the load balancer indicates the load balancing service;

The first computer device creates a local load balancer of each of the service computer devices on each of the service computer devices according to the configuration information of the load balancer to be created in each of the service computer devices;

The first computer device configures the load balancing on each of the service computer devices a forwarding mode of the service, the forwarding mode of the load balancing service initiated by the virtual machine that initiates the service on each of the service computer devices is forwarded to a local load balancer, so that the The local load balancer on each of the service computer devices performs load balancing processing on the service packets initiated by the virtual machine that initiates the service on the service computer device.
The method according to claim 9, wherein the forwarding mode of the load balancing service configured by the first computer device on each of the service computer devices comprises:

The first computer device configures the virtual switch deployed on each of the service computer devices to send the MAC address of the local load balancer to the local virtual machine that initiates the service.
The method of claim 10, wherein the first computer device configures a virtual switch deployed on each of the service computer devices to transmit the MAC address of the local load balancer to the local After the virtual machine that initiated the business, it also includes:

The virtual machine that initiates the service on each of the service computer devices sends a control packet to the local virtual switch, and the control packet carries the VIP;

The local virtual switch on each of the service computer devices sends a response message of the control packet to the local virtual machine that initiates the service, where the response message carries the MAC address of the local load balancer.
The method according to any one of claims 9 to 11, wherein the first computer device further comprises: before creating the load balancer on each of the service computer devices according to the configuration information, further comprising:

The first computer device learns that the first virtual machine of the first service computer device is created on the first service computer device, and the load balancer of the first service computer device is created on the first service computer device until A local load balancer on each of the business computer devices is created on each business computer device.
The method according to any one of claims 9 to 12, wherein the load balancer comprises a load balancing virtual machine, a load balancing container or a load balancing namespace.
A method for implementing load balancing is applied to a computer system including at least two computer devices, wherein each computer device includes a central processing unit and a memory, and the at least two computer devices include a first computer device and a second computer The device, the first computer device is a cloud management platform, the second computer device is running a virtual machine that initiates a service, and the service is a load balancing service, wherein

The first computer device obtains configuration information of a load balancer to be created in the second computer device, the configuration information of the load balancer includes an identifier of the load balancer, and a virtualized IP of the load balancer Address VIP, the VIP of the load balancer indicates the load balancing service;

The first computer device creates the load balancer on the second computer device according to the configuration information;

The first computer device configures a forwarding mode of the load balancing service on the second computer device, the forwarding mode causing the load balancing initiated by the virtual machine initiating the service on the second computer device The service packet of the service is forwarded to the load balancer, so that the load balancer selects the backend server that executes the service after receiving the service packet.
The method according to claim 14, wherein the first computer device configures a forwarding mode of the load balancing service on the second computer device, including:

The first computer device configures a virtual switch deployed on the second computer device to send a MAC address of the load balancer to the virtual machine that initiates the service, so that the virtual machine that initiates the service sends the load The MAC address of the equalizer is used as the service packet of the destination address.
The method according to claim 15, wherein the first computer device configures a virtual switch deployed on the second computer device to send the MAC address of the load balancer to the virtual machine that initiates the service. Also includes:

The virtual machine that initiates the service sends a control packet to the virtual switch, and the control packet Carrying the VIP;

The virtual switch sends a response message of the control packet to the virtual machine that initiates the service, where the response message carries the MAC address of the load balancer.
The method according to any one of claims 14 to 16, wherein the creating, by the first computer device, the load balancing on the second computer device according to the configuration information comprises:

After the first computer device creates the first virtual machine that initiates the load balancing service on the second computer device, the load equalizer is created.
The method of any of claims 14-17, wherein the method further comprises:

And deleting, by the first computer device, the created load balancer after the last one created by the second computer device initiates the load balancing service is deleted.
A computer system for implementing load balancing, comprising at least two computer devices, wherein each computer device includes a central processing unit and a memory, and the at least two computer devices include a first computer device and a second computer device, The first computer device is a cloud management platform, and the second computer device is running a virtual machine that initiates a service, where the service is a load balancing service, and the

The first computer device obtains configuration information of a load balancer to be created in the second computer device, the configuration information of the load balancer includes an identifier of the load balancer, and a virtualized IP of the load balancer Address VIP, the VIP of the load balancer indicates the load balancing service;

The first computer device creates the load balancer on the second computer device according to the configuration information;

The first computer device configures a forwarding mode of the load balancing service on the second computer device, the forwarding mode causing the load balancing initiated by the virtual machine initiating the service on the second computer device The service packet of the service is forwarded to the load balancer.

The virtual machine that initiates the service on the second computer device sends the service packet to the load balancer according to the forwarding mode;

After receiving the service packet, the load balancer on the second computer device selects at least one backend server to execute the service, and the backend server is configured to execute the load balancing service.
The computer system according to claim 19, wherein the configuring, by the first computer device, the forwarding mode of the load balancing service on the second computer device comprises:

The first computer device configures a virtual switch deployed on the second computer device, and sends a MAC address of the load balancer to the virtual machine that initiates the service;

The sending the service packet to the load balancer according to the forwarding mode of the virtual machine that initiates the service on the second computer device includes:

The virtual machine that sends the service sends the MAC address of the load balancer as the service packet of the destination address, and the virtual switch sends the service packet to the load according to the destination address of the service packet. Equalizer.
The computer system according to claim 20, wherein before the virtual machine that initiates the service sends the MAC address of the load balancer as the service message of the destination address, the method further includes:

The virtual machine that initiates the service sends a control packet to the virtual switch, where the control packet carries the VIP;

The virtual switch sends a response message of the control packet to the virtual machine that initiates the service, where the response message carries the MAC address of the load balancer.
The computer system according to claim 21, wherein the control message is an address resolution protocol ARP message or an internet control message protocol ICMP message.
The computer system according to claim 19, wherein the first computer device configures a forwarding mode of the load balancing service on the second computer device, including:

The first computer device instructs the load balancer to MAC ground the load balancer Sending the address to the virtual machine that initiated the service; or,

The first computer device configures a MAC address of the load balancer on the virtual machine that initiates the service;

Correspondingly, the sending, by the virtual machine of the initiating service on the second computer device, the sending the service packet to the load balancer according to the forwarding mode comprises:

The virtual machine that initiates the service sends the MAC address of the load balancer as the service packet of the destination address, and the virtual switch deployed on the second computer device sends the service according to the destination address of the service packet. The message is sent to the load balancer.
The computer system according to any one of claims 19 to 23, wherein the creating, by the first computer device, the load balancer on the second computer device according to the configuration information comprises:

After the first computer device creates the first virtual machine that initiates the load balancing service on the second computer device, the load equalizer is created.
A computer system according to any of claims 19-24, wherein:

And deleting, by the first computer device, the created load balancer after the last one created by the second computer device initiates the load balancing service is deleted.
A computer system according to any of claims 19-25, wherein:

The selected backend server directly returns the result of executing the service to the virtual machine that initiates the service; or

The selected backend server returns an execution result of the service to the load balancer, and the load balancer returns the execution result to the virtual machine that initiates the service.
A computer system for implementing load balancing, the computer system comprising at least three computer devices, wherein each computer device includes a central processing unit and a memory, and the at least three computer devices include a first computer device, the a computer device as a cloud management platform, wherein at least three computer devices other than the first computer device are used as business computer devices, and each business computer device runs a virtual machine that initiates a service, The service is a load balancing service, and is characterized in that

The first computer device respectively obtains configuration information of a load balancer to be created on each service computer device, where configuration information of the load balancer includes an identifier of the load balancer, and virtualization of the load balancer IP address VIP, the VIP of the load balancer indicates the load balancing service;

The first computer device creates a local load balancer of each of the service computer devices on each of the service computer devices according to the configuration information of the load balancer to be created in each of the service computer devices;

Configuring, by the first computer device, a forwarding mode of the load balancing service on each of the service computer devices, the forwarding mode causing the virtual machine initiated by the originating service on each of the service computer devices to initiate the The service packet of the load balancing service is forwarded to the local load balancer, so that the local load balancer on each service computer device initiates a service packet to the virtual machine that initiates the service on the service computer device. Perform load balancing processing.
The computer system according to claim 27, wherein the forwarding mode of the load balancing service configured by the first computer device on each of the service computer devices comprises:

The first computer device configures the virtual switch deployed on each of the service computer devices to send the MAC address of the local load balancer to the local virtual machine that initiates the service.
The computer system according to claim 28, wherein said first computer device configures a virtual switch deployed on said each business computer device, and transmits a MAC address of said local load balancer to a local office After the virtual machine that initiates the service, it also includes:

The virtual machine that initiates the service on each of the service computer devices sends a control packet to the local virtual switch, and the control packet carries the VIP;

The local virtual switch on each of the service computer devices sends a response message of the control packet to the local virtual machine that initiates the service, where the response message carries the local The MAC address of the load balancer.
The computer system according to any one of claims 27 to 29, wherein the first computer device further includes, before creating the load balancer on each of the service computer devices according to the configuration information, :

The first computer device learns that the first virtual machine of the first service computer device is created on the first service computer device, and the load balancer of the first service computer device is created on the first service computer device until A local load balancer on each of the business computer devices is created on each business computer device.
A computer system according to any one of claims 27-30, wherein the load balancer comprises a load balancing virtual machine, a load balancing container or a load balancing namespace.
A computer system for implementing load balancing, the computer system comprising at least two computer devices, wherein each computer device comprises a central processing unit and a memory, and the at least two computer devices comprise a first computer device and a second computer The device, the first computer device is a cloud management platform, the second computer device is running a virtual machine that initiates a service, and the service is a load balancing service, wherein

The first computer device obtains configuration information of a load balancer to be created in the second computer device, the configuration information of the load balancer includes an identifier of the load balancer, and a virtualized IP of the load balancer Address VIP, the VIP of the load balancer indicates the load balancing service;

The first computer device creates the load balancer on the second computer device according to the configuration information;

The first computer device configures a forwarding mode of the load balancing service on the second computer device, the forwarding mode causing the load balancing initiated by the virtual machine initiating the service on the second computer device The service packet of the service is forwarded to the load balancer, so that the load balancer selects the backend server that executes the service after receiving the service packet.
The computer system according to claim 32, wherein the first computer device configures a forwarding mode of the load balancing service on the second computer device, including:

The first computer device configures a virtual switch deployed on the second computer device to send a MAC address of the load balancer to the virtual machine that initiates the service, so that the virtual machine that initiates the service sends the load The MAC address of the equalizer is used as the service packet of the destination address.
The computer system according to claim 32, wherein said first computer device configures a virtual switch deployed on said second computer device to transmit a MAC address of said load balancer to said virtual machine that initiates service It also includes:

The virtual machine that initiates the service sends a control packet to the virtual switch, where the control packet carries the VIP;

The virtual switch sends a response message of the control packet to the virtual machine that initiates the service, where the response message carries the MAC address of the load balancer.
The computer system according to any one of claims 32-34, wherein the creating, by the first computer device, the load balancing on the second computer device according to the configuration information comprises:

After the first computer device creates the first virtual machine that initiates the load balancing service on the second computer device, the load equalizer is created.
The computer system according to any one of claims 32 to 35, wherein the method further comprises:

And deleting, by the first computer device, the created load balancer after the last one created by the second computer device initiates the load balancing service is deleted.